CN109922310A - The monitoring method of target object, apparatus and system - Google Patents

The monitoring method of target object, apparatus and system Download PDF

Info

Publication number
CN109922310A
CN109922310A CN201910069792.0A CN201910069792A CN109922310A CN 109922310 A CN109922310 A CN 109922310A CN 201910069792 A CN201910069792 A CN 201910069792A CN 109922310 A CN109922310 A CN 109922310A
Authority
CN
China
Prior art keywords
target
monitoring
image
video
local server
Prior art date
Application number
CN201910069792.0A
Other languages
Chinese (zh)
Other versions
CN109922310B (en
Inventor
臧云波
支建壮
鲁邹尧
吴明辉
Original Assignee
北京明略软件系统有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京明略软件系统有限公司 filed Critical 北京明略软件系统有限公司
Priority to CN201910069792.0A priority Critical patent/CN109922310B/en
Publication of CN109922310A publication Critical patent/CN109922310A/en
Application granted granted Critical
Publication of CN109922310B publication Critical patent/CN109922310B/en

Links

Abstract

The present invention provides a kind of monitoring methods of target object, apparatus and system, wherein, this method comprises: local server receives the target image that target monitoring equipment is sent by local net network, wherein, the target image is the image that the target monitoring equipment is shot in the case where the monitored picture for detecting the target monitoring equipment changes;The local server carries out recongnition of objects to the target image;In the case where including the target object in identifying the target image, the local server determines target identification result, wherein the target identification result is for indicating the target object occur in the monitoring area of the target monitoring equipment.Through the invention, the monitoring method for solving target object in the related technology can not stablize the lower problem of recognition efficiency of execution and target object.

Description

The monitoring method of target object, apparatus and system

Technical field

The present invention relates to the communications fields, monitoring method, apparatus and system in particular to a kind of target object.

Background technique

Currently, the monitoring method of target object in the related technology, can assist monitoring objective place using monitoring device Interior target object.For aforesaid way, the video data for shooting monitoring device is needed to be uploaded to destination server, by target Server carries out recongnition of objects.

However, the monitoring method of existing target object can not be stablized for the place of no network or unstable networks It executes, and the recognition efficiency of target object is lower.

For above-mentioned problem, currently no effective solution has been proposed.

Summary of the invention

The embodiment of the invention provides a kind of monitoring methods of target object, apparatus and system, at least to solve related skill The monitoring method of target object in art can not stablize the lower problem of recognition efficiency of execution and target object.

According to one embodiment of present invention, a kind of monitoring method of target object is provided, comprising: local server connects Receive the target image that target monitoring equipment is sent by local net network, wherein the target image is the target monitoring The image that equipment is shot in the case where the monitored picture for detecting the target monitoring equipment changes;The local service Device carries out recongnition of objects to the target image;It include the feelings of the target object in identifying the target image Under condition, the local server determines target identification result, wherein the target identification result is for indicating in the target Occurs the target object in the monitoring area of monitoring device.

Optionally, after the local server determines the target identification result, the method also includes: it is described Local server obtains the target video comprising the target image, wherein the target video includes that the target monitoring is set The standby video data shot in the target time period, the target time section are the period at the first moment to the second moment, institute State the first moment be the target monitoring equipment detect the monitored picture start variation at the time of, second moment is institute State target monitoring equipment detect the monitored picture terminate variation at the time of.

Optionally, the local server receives the institute that the target monitoring equipment is sent by the local net network Stating target image includes: that the local server receives the target monitoring equipment by the local net network in the mesh Mark the target image sent in the period every target time interval;It includes the target figure that the local server, which obtains, The target video of picture includes: not receive the target monitoring equipment in the time more than the target time interval to send The target image in the case where, the local server obtain include the target image the target video.

Optionally, it includes: the local that the local server, which obtains the target video comprising the target image, Server sends request message to the target monitoring equipment, wherein the request message includes the target figure for requesting The video of picture;It receives the target monitoring equipment and responds the target video that the request message is sent;The local service Video data that device is saved from local server concentration transfers the target video, wherein the sets of video data includes The video for the target monitoring equipment shooting that the target monitoring equipment uploads.

Optionally, after the local server obtains the target video comprising the target image, the side Method further include: the local server sends a warning message to output end, wherein the warning information is used to supervise the target It controls the appearance target object in the monitoring area of equipment to be alerted, the warning information includes at least one of: described Target image, the target video.

Optionally, before the local server sends the warning information to the output end, the method is also wrapped Include: the local server determines target information corresponding with the target video according to the target video, wherein described Target information includes at least one of: travel path of the target object in the monitoring area, the target object The invasion point into the monitoring area, the target object concealing a little in the monitoring area, the target object Activity time in the monitoring area, the characteristics of objects of the target object, the warning information further include: the target Information.

Optionally, after the local server determines the target information according to the target video, the side Method further include: the target object and the target letter corresponding with the target video are marked in the target video Breath, wherein described in the case where the target information includes travel path of the target object in the monitoring area Travel path passes through a string of punctuate marks with number in the target object on the motion track in the monitoring area Show, the number on the punctuate is sequentially increased along the moving direction of the target object.

Optionally, target corresponding with the target video is determined according to the target video in the local server After information, the method also includes: it include traveling of the target object in the monitoring area in the target information In the case where path, the local server generates prompt information according to the travel path, wherein the prompt information is used for The mode of the target object is eliminated in prompt, and the warning information further includes the prompt information.

Optionally, before the local server sends the warning information to the output end, the method is also wrapped Include: the local server obtains the objective result of multiple monitor videos, wherein the multiple monitor video is in target area Multiple monitoring devices upload monitor video, the multiple monitoring device includes the target monitoring equipment, the target knot Fruit is the target information corresponding with the multiple monitor video;According to the objective result of acquisition, the mesh is determined Mark one or more location informations in region, wherein one or more of location informations include at least one of: described Target object enters the invasion point of the target area, the target object concealing a little in the target area, the mesh Marking object frequency of occurrence in the target area is more than targeted number threshold value and/or time of occurrence more than object time threshold value Location point.

Optionally, it includes: to be in the target image that local server, which carries out recongnition of objects to the target image, In the case where multiple target video frame images, target object detection is carried out to each target video frame image, is obtained each The characteristics of image of the target video frame image, wherein each target video frame image is used to indicate in the monitored space There is the object of movement in domain, described image feature is used to indicate in the object that there is movement, with the target object Between similarity be greater than first threshold object where object region;According to each target video frame image Characteristics of image determines motion feature, wherein the motion feature is for indicating described in the multiple target video frame image There are the movement velocity of the object of movement and the directions of motion;According to the motion feature and each target video frame image Characteristics of image determines in the multiple target video frame image whether have the target object.

Optionally, target object detection is being carried out to each target video frame image, is obtaining each target view Before the described image feature of frequency frame image, the method also includes: the target monitoring equipment is obtained to the monitoring area Shoot obtained video file;Pumping frame sampling is carried out to the video file, obtains one group of video frame images;According to described one group The pixel value of pixel in video frame images determines multiple target video frame figures in one group of video frame images Picture.

Optionally, the motion feature packet is determined according to the described image feature of each target video frame image It includes: obtaining target vector corresponding with object region represented by the characteristics of image of each target video frame image, Obtain multiple target vectors, wherein each target vector is for indicating in a corresponding target video frame image Movement velocity and the direction of motion of the object that there is movement when by the object region;By the multiple target Vector forms first object vector according to time sequencing of each target video frame image in the video file, In, the motion feature includes the first object vector;Alternatively, obtaining special with the image of each target video frame image The corresponding two-dimentional light stream figure of the represented object region of sign, obtains multiple two-dimentional light stream figures, wherein each two-dimentional light Flow graph includes that the object that there is movement described in a corresponding target video frame image is passing through the object-image region Movement velocity and the direction of motion when domain;By the multiple two-dimentional light stream figure according to each target video frame image described Time sequencing in video file forms three-dimensional second object vector, wherein the motion feature includes three-dimensional second mesh Mark vector.

Optionally, according to the described image feature of the motion feature and each target video frame image, institute is determined Stating and whether having the target object in multiple target video frame images includes: by the motion feature and each target The characteristics of image of video frame images is input in preparatory trained neural network model, obtains Object identifying result, wherein institute Object identifying result is stated for indicating whether have the target object in the multiple target video frame image.

Optionally, the characteristics of image of the motion feature and each target video frame image is input to preparatory training In good neural network model, obtaining Object identifying result includes: to pass through each described image feature including convolutional layer, canonical The neural net layer structure for changing layer and activation primitive layer, obtains multiple first eigenvectors;By the multiple first eigenvector It is merged with the motion feature, obtains second feature vector;The second feature vector is input to full articulamentum to carry out Classification, obtains the first classification results, wherein the neural network model includes the neural net layer structure and the full connection Layer, the Object identifying result include first classification results, and first classification results are for indicating the multiple target Whether the target object is had in video frame images;Alternatively, it includes convolutional layer, canonical that each described image feature, which is passed through, The first nerves network layer structure for changing layer and activation primitive layer, obtains multiple first eigenvectors;The motion feature is passed through Nervus opticus network layer structure including convolutional layer, regularization layer, activation primitive layer, obtains second feature vector;It will be described more A first eigenvector is merged with the second feature vector, obtains third feature vector;By the third feature vector It is input to full articulamentum to classify, obtains the second classification results, wherein the neural network model includes the first nerves Network layer structure, the nervus opticus network layer structure and the full articulamentum, the Object identifying result include described second Classification results, second classification results are for indicating whether have the target pair in the multiple target video frame image As.

Optionally, after the local server carries out recongnition of objects to the target image, the method is also It include: in the case where not including the target object in identifying the target image, described in the local server deletion The target video that target monitoring equipment uploads, wherein the target video includes the target monitoring equipment in target time section The video data of interior shooting, the target time section are the period at the first moment to the second moment, and first moment is institute State target monitoring equipment detect the monitored picture start variation at the time of, second moment be the target monitoring equipment At the time of detecting that the monitored picture terminates variation.

According to another embodiment of the invention, a kind of monitoring method of target object is provided, comprising: monitoring device pair The monitoring area of the monitoring device carries out mobile detection, wherein the movement detects the prison for detecting the monitoring device Whether control picture changes;In the case where detecting that the monitored picture changes, the monitoring device is obtained to institute State the target image that monitoring area is shot;The monitoring device uploads the target image by local net network To local server, wherein the target image is used to indicate the local server and carries out target pair to the target image As identification.

Optionally, the target image is uploaded to the local service by the local net network by the monitoring device Device includes: the mesh that the monitoring device takes the monitoring device every target time interval in the target time period Logo image is uploaded to the local server, wherein the target time section is the period at the first moment to the second moment, institute State the first moment be the monitoring device detect the monitored picture start variation at the time of, second moment be the prison At the time of control equipment detects that the monitored picture terminates variation.

Optionally, the target image is uploaded to the local clothes by the local net network in the monitoring device It is engaged in after device, the method also includes: in the case where detecting that the monitored picture terminates variation, the monitoring device is by mesh Mark video is uploaded to the local server, wherein the target video includes that the monitoring device is clapped in the target time period The video data taken the photograph, the target time section are the period at the first moment to the second moment, and first moment is the prison At the time of control equipment detects that the monitored picture starts variation, second moment is that the monitoring device detects the prison At the time of control picture terminates variation.

Optionally, before target video is uploaded to the local server by the monitoring device, the method is also wrapped Include: the monitoring device determines first moment and second moment;The monitoring device is from the monitoring device The target video is read out in memory, wherein the target video is the third moment of monitoring device shooting to the 4th The video data at moment, before the third moment is first moment and first moment is separated by first time period At the moment, the 4th moment is after second moment, at the time of being separated by the second time interval with second moment.

Optionally, before the target video is uploaded to the local server by the monitoring device, the method is also It include: in the case where detecting that the monitored picture terminates variation, the monitoring device is from the memory of the monitoring device Read out the target video;The target video is written in the storage unit of the monitoring device by the monitoring device.

Optionally, it includes: by the mesh that the target video is uploaded to the local server by the monitoring device During mark video is uploaded to the local server, the monitoring device detects that the local net network is in abnormal shape State;The monitoring device saves the upload progress of the target video;Detecting that the local net network is in normal condition In the case where, the monitoring device continues to upload the target view to the local server according to the upload progress of preservation Frequently.

Optionally, it includes: described that the monitoring device, which carries out mobile detection to the monitoring area of the monitoring device, Monitoring device carries out mobile detection to the monitoring area of the monitoring device by the camera of the monitoring device.

According to still another embodiment of the invention, a kind of monitoring system of target object is provided, comprising: target monitoring is set Standby and local server, the target monitoring equipment are connected with the local server by local net network, wherein the mesh Monitoring device is marked, carries out mobile detection for the monitoring area to the target monitoring equipment, wherein the mobile detection is used for Whether the monitored picture for detecting the target monitoring equipment changes;Detecting the case where monitored picture changes Under, obtain the target image shot to the monitoring area;The target image is passed through into the local net network It is uploaded to the local server;The local server passes through the local area network for receiving the target monitoring equipment The target image that network is sent;Recongnition of objects is carried out to the target image;In identifying the target image In the case where including the target object, target identification result is determined, wherein the target identification result is for indicating Occurs the target object in the monitoring area.

Optionally, the target monitoring equipment is also used to: by the target monitoring equipment in the target time period every mesh The target image that mark time interval takes is uploaded to the local server, wherein the target time section is first The period at moment to the second moment, first moment are that the monitoring device detects that the monitored picture starts variation Moment, at the time of second moment is that the monitoring device detects that the monitored picture terminates variation.

Optionally, the target monitoring equipment is also used to:, will in the case where detecting that the monitored picture terminates variation Target video is uploaded to the local server, wherein the target video include the monitoring device in the target time period The video data of shooting, the target time section are the period at the first moment to the second moment, and first moment is described At the time of target monitoring equipment detects that the monitored picture starts variation, second moment is target monitoring equipment inspection At the time of measuring the monitored picture terminates variation.

Optionally, the system also includes output end, the output end is connected with the local server, wherein described Local server is also used to receive the target video that the target monitoring equipment uploads;It is determined according to the target video Target information corresponding with the target video out, wherein the target information includes at least one of: the target object Travel path in the monitoring area, the invasion point into the monitoring area of the target object, the target pair As concealing a little in the monitoring area, activity time of the target object in the monitoring area, the target pair The characteristics of objects of elephant;The case where the target information includes travel path of the target object in the monitoring area Under, prompt information is generated according to the travel path, wherein the prompt information is used to prompt to eliminate the side of the target object Formula;It sends a warning message to the output end, wherein the warning information is used for the monitoring area to the target monitoring equipment The interior appearance target object is alerted, and the warning information includes at least one of: the target image, the target Video, the target information, the prompt information;The output end, for receiving the warning information;In the output end At least one of is shown on screen: the target image, the target video, the target information, the prompt information.

Optionally, the local server is also used to not include the target object in identifying the target image In the case where, delete the target video that the target monitoring equipment uploads, wherein the target video includes the target monitoring The video data that equipment is shot in the target time period, the target time section are the period at the first moment to the second moment, At the time of first moment is that the target monitoring equipment detects that the monitored picture starts variation, second moment is At the time of the target monitoring equipment detects that the monitored picture terminates variation.

According to still another embodiment of the invention, a kind of monitoring device of target object is provided, local service is applied to Device, comprising: receiving module, the target image sent for receiving target monitoring equipment by local net network, wherein described Target image is the target monitoring equipment in the case where the monitored picture for detecting the target monitoring equipment changes The image of shooting;Identification module, for carrying out recongnition of objects to the target image;Determining module, for identifying In the case where including the target object in the target image, target identification result is determined, wherein the target identification As a result for indicating the target object occur in the monitoring area of the target monitoring equipment.

According to still another embodiment of the invention, a kind of monitoring device of target object is provided, monitoring device is applied to, Include: detection module, carry out mobile detection for the monitoring area to the monitoring device, wherein the mobile detection is used for Whether the monitored picture for detecting the monitoring device changes;Module is obtained, for detecting the monitored picture generation In the case where variation, the target image shot to the monitoring area is obtained;Uploading module is used for the target Image is uploaded to local server by local net network, wherein the target image is used to indicate the local server pair The target image carries out recongnition of objects.

According to still another embodiment of the invention, a kind of storage medium is additionally provided, meter is stored in above-mentioned storage medium Calculation machine program, wherein above-mentioned computer program is arranged to execute the step in any of the above-described embodiment of the method when operation.

According to still another embodiment of the invention, a kind of electronic device, including memory and processor are additionally provided, is stored Computer program is stored in device, above-mentioned processor is arranged to run computer program to execute the implementation of any of the above-described method Step in example.

Through the invention, local server receives the target image that target monitoring equipment is sent by local net network, Wherein, target image is that target monitoring equipment is shot in the case where the monitored picture for detecting target monitoring equipment changes Image;Local server carries out recongnition of objects to target image;It include target object in identifying target image In the case where, local server determines target identification result, wherein target identification result is for indicating in target monitoring equipment Monitoring area in there is target object, due to by local net network carry out target image upload, without broadband network Intervention, local net network is relatively stable, and avoiding the relevant operation as caused by no network or unstable networks can not stablize It executes, also, since what is transmitted by local net network is target image, the data volume of transmission is small, reduces transmission time, and Local server can determine whether monitoring area target object occurs according to the image received, improve transmission speed, in turn Monitoring efficiency is improved, therefore, the monitoring method that can solve target object in the related technology can not stablize execution and mesh Mark the lower problem of the recognition efficiency of object.

Detailed description of the invention

The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, this hair Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:

Fig. 1 is a kind of hardware block diagram of the local server of the monitoring method of target object of the embodiment of the present invention;

Fig. 2 is the flow chart one of the monitoring method of target object according to an embodiment of the present invention;

Fig. 3 is the flowchart 2 of the monitoring method of target object according to an embodiment of the present invention;

Fig. 4 is a kind of schematic diagram of each module data connection according to the preferred embodiment of the invention;

Fig. 5 is the schematic illustration that a kind of mouse according to the preferred embodiment of the invention suffers from detection method;

Fig. 6 is a kind of schematic diagram of Faster-RCNN network model of the preferred embodiment of the present invention;

Fig. 7 is the structural block diagram one of the monitoring device of target object according to an embodiment of the present invention;

Fig. 8 is the structural block diagram two of the monitoring device of target object according to an embodiment of the present invention;

Fig. 9 is the structural block diagram of the monitoring system of target object according to an embodiment of the present invention;

Figure 10 is the schematic diagram of the monitoring framework of the target object of alternative embodiment according to the present invention.

Specific embodiment

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings and in combination with Examples.It should be noted that not conflicting In the case of, the features in the embodiments and the embodiments of the present application can be combined with each other.

It should be noted that description and claims of this specification and term " first " in above-mentioned attached drawing, " Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.

Embodiment 1

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings and in combination with Examples.It should be noted that not conflicting In the case of, the features in the embodiments and the embodiments of the present application can be combined with each other.

It should be noted that description and claims of this specification and term " first " in above-mentioned attached drawing, " Two " etc. be to be used to distinguish similar objects, without being used to describe a particular order or precedence order.

Embodiment of the method provided by the embodiment of the present application 1 can be in local server, terminal or similar It is executed in arithmetic unit.For operating on local server, Fig. 1 is a kind of monitoring of target object of the embodiment of the present invention The hardware block diagram of the local server of method.As shown in Figure 1, local server 10 may include one or more (in Fig. 1 Only showing one) (processor 102 can include but is not limited to Micro-processor MCV or programmable logic device FPGA to processor 102 Deng processing unit) and memory 104 for storing data, optionally, above-mentioned local server can also include for leading to The transmission device 106 and input-output equipment 108 of telecommunication function.It will appreciated by the skilled person that knot shown in FIG. 1 Structure is only to illustrate, and does not cause to limit to the structure of above-mentioned local server.For example, local server 10 may also include than figure More or less component shown in 1, alternatively, having the configuration different from shown in Fig. 1.

Memory 104 can be used for storing computer program, for example, the software program and module of application software, such as this hair The corresponding computer program of the monitoring method of target object in bright embodiment, processor 102 are stored in memory by operation Computer program in 104 realizes above-mentioned method thereby executing various function application and data processing.Memory 104 May include high speed random access memory, may also include nonvolatile memory, as one or more magnetic storage device, flash memory, Or other non-volatile solid state memories.In some instances, memory 104 can further comprise relative to processor 102 Remotely located memory, these remote memories can pass through network connection to local server 10.The example of above-mentioned network Including but not limited to internet, intranet, local area network, mobile radio communication and combinations thereof.

Transmitting device 106 is used to that data to be received or sent via a network.Above-mentioned network specific example may include The wireless network that the communication providers of local server 10 provide.In an example, transmitting device 106 includes that a network is suitable Orchestration (Network Interface Controller, referred to as NIC), can be connected by base station with other network equipments from And it can be communicated with internet.In an example, transmitting device 106 can be radio frequency (Radio Frequency, abbreviation For RF) module, it is used to wirelessly be communicated with internet.

A kind of monitoring method of target object is provided in the present embodiment, and Fig. 2 is target according to an embodiment of the present invention The flow chart one of the monitoring method of object, as shown in Fig. 2, the process includes the following steps:

Step S202, local server receive the target image that target monitoring equipment is sent by local net network, In, target image is what target monitoring equipment was shot in the case where the monitored picture for detecting target monitoring equipment changes Image;

Step S204, local server carry out recongnition of objects to target image;

In the case that step S206 in identifying target image includes target object, local server determines mesh Identify other result, wherein target identification result is for indicating target object occur in the monitoring area of target monitoring equipment.

Optionally, in the present embodiment, target object can include but is not limited to: mouse, cockroach etc. harmful organism (that is, target organism);Monitoring area can include but is not limited to: the monitoring area in the places such as kitchen, warehouse, workshop;Target Monitoring device can include but is not limited to: camera, monitor etc.;Target monitoring equipment can include but is not limited to: one Or multiple monitoring devices.

Through the above steps, local server receives the target figure that target monitoring equipment is sent by local net network Picture, wherein target image is target monitoring equipment in the case where the monitored picture for detecting target monitoring equipment changes The image of shooting;Local server carries out recongnition of objects to target image;It include target in identifying target image In the case where object, local server determines target identification result, wherein target identification result is for indicating in target monitoring There is target object in the monitoring area of equipment, the monitoring method for solving target object in the related technology, which can not be stablized, to be held The lower problem of capable and target object recognition efficiency, ensure that the stability that monitoring executes, reduces transmission time, mention High transmission speed, and then improve monitoring efficiency.

In step S202, local server receives the target figure that target monitoring equipment is sent by local net network Picture, wherein target image is target monitoring equipment in the case where the monitored picture for detecting target monitoring equipment changes The image of shooting.

In order to realize shooting at night, target monitoring equipment can have camera.Target monitoring equipment after actuation, can be with Its monitoring area is shot always, the video of shooting is stored in memory.If memory has been expired, the monitoring newly shot is regarded Frequency can cover the video shot earliest.

Optionally, above-mentioned camera can include but is not limited to: the camera with infrared illumination function, for example, infrared Lll night vision camera.Further, which can also include but is not limited to: mobile detection function, store function, networking Function (such as wifi networking) and fine definition (e.g., being greater than 1080p) configuration.

The mode of mobile detection can be used in target monitoring equipment, when produced image content changes (that is, monitoring picture Face changes) in the case where, the image of shooting is sent to local server.The reason of monitored picture changes can wrap It includes at least one of: having mouse appearance, there is cockroach appearance, other target organisms occur, and foreign matter is flown into.

Optionally, target monitoring equipment is in the case where the monitored picture for detecting target monitoring equipment changes, to The target image that local server is sent can be one or more.

As an alternative embodiment, target image can be target prison in the case where target image is one Equipment is controlled in the case where detecting that monitored picture changes, the image of current shooting is sent to local server.

As another optional embodiment, in the case where target image is multiple, target image can be target Monitoring device saves the image of a shooting in the case where detecting that monitored picture changes, every target time interval, And picture is sent to local server in real time.

Target monitoring equipment can start to be changed to the mesh for detecting that monitored picture terminates variation detecting monitored picture It marks in the period, periodically sends target image to local server.One after detecting that monitored picture terminates variation It fixes time, target monitoring equipment periodically can also send target image to local server, so that local server is to mesh Logo image is analyzed.

Optionally, local server receives target monitoring equipment and can wrap by the target image that local net network is sent Include: local server receives target monitoring equipment and is sent out in the target time period every target time interval by local net network The target image sent.

In step S204, local server carries out recongnition of objects to target image.

Local server can be completed to the image recognition of target image after receiving target image (for target The image recognition of object), using AI technology, judge whether there is target object (e.g., target organism) appearance in image, or only Foreign matter such as flies at the non-insect pest invasion scene.

When carrying out recongnition of objects, the object of identification can be one or more target images.In the object of identification In the case where multiple target images, local server can be comprehensive each in conjunction with the characteristics of image for including in multiple target images Target image obtains recognition result.

Optionally, it is multiple that local server, which may include: in target image to target image progress recongnition of objects, In the case where target video frame image, target object detection is carried out to each target video frame image, obtains each target video The characteristics of image of frame image, wherein each target video frame image is used to indicate the object that there is movement in monitoring area, figure As feature is used to indicate that the similarity in the object in the presence of movement, between target object to be greater than where the object of first threshold Object region;Motion feature is determined according to the characteristics of image of each target video frame image, wherein motion feature is used In the movement velocity and the direction of motion of the object for indicating to exist in multiple target video frame images movement;According to motion feature and often The characteristics of image of a target video frame image determines in multiple target video frame images whether have target object.It is above-mentioned each Step can be executed by local server.

Optionally, target object detection is being carried out to each target video frame image, is obtaining each target video frame image Characteristics of image before, video file that available target monitoring equipment shoots monitoring area;To video file into Row takes out frame sampling, obtains one group of video frame images;According to the pixel value of the pixel in one group of video frame images in one group of video Multiple target video frame images are determined in frame image.

The step of above-mentioned acquisition video file, sampling, can be executed by target monitoring equipment, the above-mentioned multiple target views of determination The step of frequency frame image, can be executed by target monitoring equipment, can also be executed by local server.

As an alternative embodiment, determining motion feature according to the characteristics of image of each target video frame image It may include: to obtain target arrow corresponding with object region represented by the characteristics of image of each target video frame image Amount, obtains multiple target vectors, wherein each target vector is for indicating there is fortune in corresponding target video frame image Movement velocity and the direction of motion of the dynamic object when by object region;Multiple target vectors are regarded according to each target Time sequencing of the frequency frame image in video file forms first object vector, wherein motion feature includes first object vector.

As another optional embodiment, determine that movement is special according to the characteristics of image of each target video frame image Sign may include: to obtain two-dimentional light corresponding with object region represented by the characteristics of image of each target video frame image Flow graph obtains multiple two-dimentional light stream figures, wherein each two dimension light stream figure includes existing in corresponding target video frame image Movement velocity and the direction of motion of the object of movement when by object region;By multiple two-dimentional light stream figures according to each mesh It marks time sequencing of the video frame images in video file and forms three-dimensional second object vector, wherein motion feature includes three-dimensional Second object vector.

Optionally, according to the characteristics of image of motion feature and each target video frame image, multiple target video frames are determined It may include: to input the characteristics of image of motion feature and each target video frame image that target object whether is had in image Into preparatory trained neural network model, Object identifying result is obtained, wherein Object identifying result is for indicating multiple mesh Whether target object is had in mark video frame images.

As an alternative embodiment, the characteristics of image of motion feature and each target video frame image is input to In preparatory trained neural network model, obtaining Object identifying result may include: to pass through each characteristics of image including volume The neural net layer structure of lamination, regularization layer and activation primitive layer, obtains multiple first eigenvectors;By multiple fisrt feature Vector is merged with motion feature, obtains second feature vector;Second feature vector is input to full articulamentum to classify, Obtain the first classification results, wherein neural network model includes neural net layer structure and full articulamentum, Object identifying result packet The first classification results are included, the first classification results are for indicating whether have target object in multiple target video frame images.

As an alternative embodiment, the characteristics of image of motion feature and each target video frame image is input to In preparatory trained neural network model, obtaining Object identifying result may include: to pass through each characteristics of image including volume The first nerves network layer structure of lamination, regularization layer and activation primitive layer, obtains multiple first eigenvectors;By motion feature By including the nervus opticus network layer structure of convolutional layer, regularization layer, activation primitive layer, second feature vector is obtained;It will be more A first eigenvector is merged with second feature vector, obtains third feature vector;Third feature vector is input to entirely Articulamentum is classified, and the second classification results are obtained, wherein neural network model includes first nerves network layer structure, second Neural net layer structure and full articulamentum, Object identifying result include the second classification results, and the second classification results are for indicating more Whether target object is had in a target video frame image.

In step S206, in the case where in identifying target image including target object, local server is determined Target identification result out, wherein target identification result is for indicating target pair occur in the monitoring area of target monitoring equipment As.

The result for carrying out recongnition of objects to target image can include but is not limited to following two: target image includes There is target object (that is, target identification result);Target image does not include target object.

Optionally, after local server determines target identification result, available local server includes target The target video of image, wherein target video includes the video data that target monitoring equipment is shot in the target time period, target Period is the period at the first moment to the second moment, and the first moment was that target monitoring equipment detects that monitored picture starts to become At the time of change, at the time of the second moment was that target monitoring equipment detects that monitored picture terminates variation.

The mode that local server obtains target video can include but is not limited to following one:

Mode one, local server obtain target video from target monitoring equipment.

Local server can send request message to target monitoring equipment, wherein request message includes mesh for requesting The video of logo image;Receive the target video that target monitoring equipment response request message is sent

Mode two, the sets of video data that local server is saved from local server obtain target video.

The video data concentration that local server can be saved from local server transfers target video, wherein video counts According to the video for the target monitoring equipment shooting that collection is uploaded comprising target monitoring equipment.

In the case where detecting that the monitored picture of target monitoring equipment changes, in addition to the target image hair that will acquire It gives except local server, target monitoring equipment can also include the video counts shot in target time section from obtaining in memory According to (that is, target video).

Optionally, target monitoring equipment can be in the case where detecting that monitored picture terminates variation, for periodical hair The scene of target image is sent, target monitoring equipment stops sending target image to local server, and obtains the view in the period Frequently.Can pre-record and be delayed certain time relative to target time section the above-mentioned period.That is, the above-mentioned period are as follows: the third moment to Time interval between four moment, before the third moment was the first moment and at the time of the first moment was separated by first time period, the Four moment is after the second moment, at the time of being separated by the second time interval with the second moment.

For example, target monitoring equipment detects that monitored picture changes in 01:00:00, prison is detected in 01:05:00 Control picture variation terminates, 2 seconds before the available 01:00:00 of target monitoring equipment after 2 seconds to 01:05:00 (00: The video of 59:58 to 01:05:02).

It should be noted that a buffer time can be set to guarantee to determine the accuracy that monitored picture terminates variation (for example, 5s), to avoid being persisted due to object (e.g., target object) in a certain position caused by monitored picture do not change.

Target monitoring equipment obtain target video opportunity can be receive local server transmission above-mentioned request disappear Breath can also voluntarily be determined when to obtain by target monitoring equipment target video (for example, detecting that monitored picture terminates to change Later).

After getting target video, target video can be stored in target monitoring equipment by target monitoring equipment In storage unit (for example, being inserted into the SD card in target monitoring equipment).

The mode that local server obtains target video can include but is not limited at least one of:

Local server sends request message to target monitoring equipment;Target monitoring equipment obtains target view from memory Frequently, alternatively, obtaining target video from the video stored in storage unit;The target video that will acquire is sent to local service Device;

Target monitoring equipment is after getting target video in memory, alternatively, target video is deposited into storage unit After part, by target video active transmission to local server.

The opportunity that local server sends request message can be at least one of: detecting in target image and includes After target object (in the case where target image is one);Time more than target time interval does not receive target monitoring The target image that equipment is sent (in the case where target image is multiple, send out target monitoring equipment in the target time period by the period Send target image).

Target monitoring equipment can have broken string and resume function, when wireless network environment is unstable, can also guarantee to regard Frequency is completely uploaded to local server (resume or upload again) later.

Optionally, it may include: will be on target video that target video is uploaded to local server by target monitoring equipment During reaching local server, detect that local net network is in abnormality;Save the upload progress of target video;? In the case where detecting that local net network is in normal condition, continued to upload mesh to local server according to the upload progress of preservation Mark video.

Above-mentioned abnormality can include but is not limited to: without network, unstable networks etc..

Optionally, after local server obtains the target video comprising target image, local server can be to defeated Outlet sends a warning message, wherein warning information is used for there is target object progress in the monitoring area of target monitoring equipment Alarm, warning information includes at least one of: target image, target video.

Before local server sends a warning message to output end, local server can be determined according to target video Target information corresponding with target video, wherein target information includes at least one of: target object is in monitoring area Travel path, the invasion point into monitoring area of target object, target object concealing a little in monitoring area, target object Activity time in monitoring area, the characteristics of objects of target object, warning information further include: target information.

After getting target video, local server can analyze corresponding with target video from target video Target information.The target information of acquisition is for the dominant target object that indicates in target video (for example, pair of target object As feature), the action message of target object is (for example, travel path of the target object in monitoring area, the entrance of target object The invasion point of monitoring area, target object concealing a little in monitoring area, activity time of the target object in monitoring area) Deng.The object information of target object can include but is not limited to: the type of target object, the skin color etc. of target object.

Optionally, after local server determines target information according to target video, local server can be in mesh Target object and target information corresponding with target video are marked in mark video.

Target information can be marked in target video in a manner of information list, can also in target video with target Target information is represented on the position of information association.

For example, the characteristics of objects of target object can be marked in a manner of information list, marked in target video: mesh The travel path of object, the invasion point of target object are marked, target object is concealed a little, the activity time of target object.

In the case where target information includes the conduct path of target object, travel path in target object by monitoring A string on motion track in region have the punctuate mark of number, the number on punctuate along target object moving direction according to Secondary increase.

For example, travel path, which is shown, can be used punctuate expression, with a string of digital representations from small to large on line segment, with table Show the direction of travel of mouse or cockroach.

Optionally, local server can generate according to the travel track of the target object analyzed and eliminate target object It is recommended that being supplied to user.Such as: traveling of the target object in monitoring area is determined in target video in local server After track (e.g., motion track), local server generates prompt information according to travel track, wherein prompt information is for mentioning Show the mode for eliminating target object.

After generating prompt information, prompt information can individually be sent to output end, can also be with other information (e.g., Target image, target video, target information etc.) send (e.g., being sent by warning information) together.

If there is multiple monitoring devices in target place, local server can be to the monitoring of multiple monitoring devices upload Data are merged, so that it is determined that out in entire target place target object characteristics of objects and movement routine etc..

Optionally, local server obtains the objective result of multiple monitor videos, wherein multiple monitor videos are target area The monitor video that multiple monitoring devices in domain upload, multiple monitoring devices include target monitoring equipment, objective result be with it is more The corresponding target information of a monitor video;According to the objective result of acquisition, one or more positions in target area are determined Information, wherein one or more location informations include at least one of: target object enters the invasion point of target area, mesh Mark object concealing a little in target area, target object in target area frequency of occurrence be more than targeted number threshold value and/or Time of occurrence is more than the location point of object time threshold value.

According to local server to the analysis of multiple monitoring datas as a result, can determine crucial location point (invasion point, Conceal a little and the location point of frequent activity etc.), to provide treatment advice according to the key position point determined.

For example, local server can summarize the worm plague of rats information that multiple monitoring devices are collected, haunted according to the worm plague of rats Historical path is suitble to dispose the position of the instruments such as mouse sticking plate, cockroach-killing room in conjunction with place place, and the position for providing placement is suggested.

Warning information may include at least one of: target image, target video, target information (target video and mesh Mark information can be combined together display) and treatment advice.

After local server carries out recongnition of objects to target image, local server can be in several ways Storage resource is saved, unnecessary data transmission or data storage are reduced.

It is determining that the upload of target complete image is completed in target monitoring equipment, and is not detecting mesh in the target image In the case where marking object, local server can execute following operation:

Instruction information is sent to target monitoring equipment, which can serve to indicate that target monitoring equipment not by target Video is saved in storage unit, alternatively, target video is not sent to local server;Alternatively,

Delete the target video locally saved.

The scene that a target image is only uploaded for target monitoring equipment, after receiving target image, local service Device can determine that the upload of target complete image is completed in target monitoring equipment.

The scene that multiple target images are uploaded for target monitoring equipment does not receive in the time more than target time interval When the target image sent to target monitoring equipment, alternatively, receive the transmission of target monitoring equipment be used to indicate be completed it is complete When the instruction information of the upload of portion's target image, local server can determine that target complete image is completed in target monitoring equipment Upload.

The monitoring method of another target object is additionally provided in the present embodiment, and Fig. 3 is according to an embodiment of the present invention The flowchart 2 of the monitoring method of target object, as shown in figure 3, the process includes the following steps:

Step S302, monitoring device carry out mobile detection to the monitoring area of monitoring device, wherein mobile detection is for examining Whether the monitored picture for surveying monitoring device changes;

Step S304, in the case where detecting that monitored picture changes, monitoring device, which obtains, carries out monitoring area Shoot obtained target image;

Target image is uploaded to local server by local net network by step S306, monitoring device, wherein target figure Recongnition of objects is carried out to target image as being used to indicate local server.

Optionally, in the present embodiment, the function of monitoring device and preceding aim monitoring device, execution operation can phase Together.Target object, monitoring area, monitoring device with it is aforementioned substantially similar, this will not be repeated here.

Through the above steps, monitoring device carries out mobile detection to the monitoring area of monitoring device, wherein mobile detection is used Whether change in the monitored picture of detection monitoring device;In the case where detecting that monitored picture changes, monitoring is set It is standby to obtain the target image shot to monitoring area;Target image is uploaded to by monitoring device by local net network Local server, wherein target image is used to indicate local server and carries out recongnition of objects to target image, solves phase The monitoring method of target object in the technology of pass can not stablize the lower problem of recognition efficiency of execution and target object, protect The stability that monitoring executes has been demonstrate,proved, has reduced transmission time, improves transmission speed, and then improve monitoring efficiency.

Optionally, it includes: monitoring device that target image is uploaded to local server by local net network by monitoring device Monitoring device is uploaded to local server every the target image that target time interval takes in the target time period, In, target time section is the period at the first moment to the second moment, and the first moment was that monitoring device detects that monitored picture is opened At the time of beginning to change, at the time of the second moment was that monitoring device detects that monitored picture terminates variation.

Optionally, it after target image is uploaded to local server by local net network by monitoring device, is detecting In the case where terminating variation to monitored picture, target video can be uploaded to local server by monitoring device, wherein target view The video data that frequency is shot in the target time period comprising monitoring device, target time section be the first moment to the second moment when Between section, the first moment be monitoring device detect monitored picture start variation at the time of, the second moment detected for monitoring device At the time of monitored picture terminates variation.

Optionally, before target video is uploaded to local server by monitoring device, can determine the first moment and Second moment;Target video is read out from the memory of monitoring device, wherein when target video is the third of monitoring device shooting It carves to the video data at the 4th moment, before the third moment was the first moment and at the time of the first moment was separated by first time period, 4th moment is after the second moment, at the time of being separated by the second time interval with the second moment.

Optionally, before target video is uploaded to local server by monitoring device, detecting that monitored picture terminates In the case where variation, monitoring device reads out target video from the memory of monitoring device;Target video is written to monitoring to set In standby storage unit.

Optionally, it includes: that target video is being uploaded to local that target video is uploaded to local server by monitoring device During server, detect that local net network is in abnormality;Save the upload progress of target video;Detecting office In the case that domain net network is in normal condition, continued to upload target video to local server according to the upload progress of preservation.

Optionally, it includes: that monitoring device passes through monitoring that monitoring device, which carries out mobile detection to the monitoring area of monitoring device, The camera of equipment carries out mobile detection to the monitoring area of monitoring device.

A kind of determination method of target object is additionally provided in the present embodiment.The determination method of above-mentioned target object can be with It is applied in the monitoring method of above-mentioned target object.Assuming that target monitoring equipment is picture pick-up device.Optionally, as a kind of optional Embodiment, the determination method of above-mentioned target object may comprise steps of:

Step S1 obtains the video file that picture pick-up device shoots monitoring area.

Picture pick-up device can be monitoring camera, for example, the picture pick-up device is infrared low-light night vision camera, for it Monitoring area is shot, and video file is obtained.Wherein, monitoring area is to be detected region, that is, to have detected whether target The region that object occurs, which can be the biggish Vector prevented and treated of figure, for example, the target pair As for mouse.

The video file may include the original video data shot to monitoring area, may include monitored space The monitor video sequence in domain, the monitor video sequence namely video stream sequence.

It is alternatively possible to the original video data of monitoring area be obtained by ARM plate in video acquisition module, on generating Video file is stated, to realize the purpose being acquired to the video of monitoring area.

Step S2 carries out pumping frame sampling to video file, obtains one group of video frame images.

After obtaining the video file that picture pick-up device shoots monitoring area, video file is pre-processed, Pumping frame sampling can be carried out to video file in video data process layer, obtain one group of video frame images.

Equally spaced pumping frame sampling can be carried out to video file, so that one group of video frame images of video file are obtained, For example, video file includes that 100 sequence of frames of video obtain 10 sequence of frames of video, then by this after carrying out pumping frame sampling 10 sequence of frames of video are as above-mentioned one group of video frame images, to reduce the fortune of the algorithm for being determined to target object Calculation amount.

Step S3 is determined in one group of video frame images more according to the pixel value of the pixel in one group of video frame images A target video frame image.

Pumping frame sampling is being carried out to video file, it, can be according to one group of video frame figure after obtaining one group of video frame images The pixel value of pixel as in determines multiple target video frame images in one group of video frame images, wherein each target Video frame images are used to indicate the object that there is movement in corresponding monitoring area.

Video file can be pre-processed, further include that dynamic detection is carried out to video file, from one group of video frame figure The target video frame image for being used to indicate the object that there is movement in monitoring area is determined as in, that is, in the target video There is the object of movement in frame image, which can be the video clip of the object in the presence of movement, wherein deposit It may be target object in the object of movement, it is also possible to not be.The embodiment can determine that target is regarded by dynamic detection algorithm Frequency frame image determines multiple targets in one group of video frame images according to the pixel value of the pixel in one group of video frame images Video frame images, and then execute step S4.

Optionally, in one group of video frame images, the video frame images in addition to multiple target video frame images are not indicated There is the image of movement in corresponding monitoring area out, it can be without subsequent detection.

Step S4 carries out target object detection to each target video frame image, obtains each target video frame image Characteristics of image.

Multiple mesh are being determined in one group of video frame images according to the pixel value of the pixel in one group of video frame images After marking video frame images, target object detection is carried out to each target video frame image, obtains each target video frame image Characteristics of image, wherein characteristics of image is deposited in target video frame image for each target video frame image for indicating When the object of movement is judged as target object, there are the object regions where the object of movement.

Target object detection is carried out to each target video frame image, that is, transporting to present in target video frame image Dynamic object is detected, and can be examined by object detection system using Detection dynamic target method and target neural network based Survey method detects Moving Objects present in target video frame image, and the image for obtaining each target video frame image is special Sign, wherein the arithmetic speed of Detection dynamic target method is fast, lower to machine configuration requirement, and target neural network based The accuracy and robustness of detection method are more preferable, and characteristics of image can be the visual information in rectangle frame, for indicating target figure As region, which can be detection block, for indicating in there is the object moved, with the target object to be identified it Between similarity be greater than first threshold object where object region.That is, above-mentioned characteristics of image is used to indicate The position that the target object that scalping is confirmed is likely to occur.

Step S5 determines motion feature according to the characteristics of image of each target video frame image.

Target object detection is being carried out to each target video frame image, the image for obtaining each target video frame image is special After sign, motion feature is determined according to the characteristics of image of each target video frame image, wherein motion feature is for indicating more There is the movement velocity and the direction of motion of the object of movement in a target video frame image.

Target object detection is being carried out to each target video frame image, the image for obtaining each target video frame image is special After sign, the characteristics of image of each target video frame image can be input to motion feature extraction module, which mentions Modulus root tuber determines motion feature according to the characteristics of image of each target video frame image, which regards for multiple targets For frequency frame image, the movement velocity and the direction of motion of the object for indicating to exist in multiple target video frame images movement, Interference image caused by the movement of non-targeted object is further filtered out simultaneously, for example, deleting the mobile of mosquito waits interference Information.

It optionally, is continuous, motion feature due to there is the movement of the object of movement in each target video frame image The motion feature extraction algorithm of extraction module can be first according to the multiple targets of Image Feature Detection of each target video frame image The corresponding object of the big characteristics of image of correlation can be determined as same by the correlation of the characteristics of image between video frame images Object matches the characteristics of image of each target video frame image, obtains the range of motion picture of object, finally can be with The feature that motion sequence is extracted using the feature extraction network of 3D, so that motion feature is obtained, for example, according to each target video The detection block of frame image calculates the correlation of detection block between multiple target video frame images, detection that can be big by correlation The corresponding object of frame is determined as same target, matches to the detection block of each target video frame image, obtains the one of object Movement sequence picture finally extracts the feature of motion sequence using the feature extraction network of 3D, obtains motion feature, and then determine There is the movement velocity and the direction of motion of the object of movement in multiple target video frame images.

Alternatively it is also possible to by the characteristics of image of multiple target video frame images carry out fusion and and carry out feature extraction, The case where to prevent the object detector of single frames from judging by accident, and then realize and fine screen is carried out accurately to determine to target image Whether target object is occurred.

Step S6 determines multiple target video frames according to the characteristics of image of motion feature and each target video frame image Whether target object is had in image.

After determining motion feature according to the characteristics of image of each target video frame image, can by motion feature and The characteristics of image of each target video frame image is merged, and is input in preparatory trained sorter network, the sorter network To be pre-designed for determining the sorter network model for whether having target object in multiple target video frame images, into And according to the characteristics of image of motion feature and each target video frame image, determine whether occur in multiple target video frame images There is target object, for example, determining in multiple target video frame images whether have mouse.

It is alternatively possible to which the characteristics of image for the target video frame for having target object in multiple target video frame images is inputted To front end (output end) display interface, which can show the detection block and moving rail of target object in turn Mark.

Optionally, the sorter network model of the embodiment can be used for filtering the sequence of pictures of non-targeted object, and retain The sequence of pictures of target object guarantees the accuracy of target object prompt information to reduce false alarm rate.

In the present embodiment, video file monitoring area shot by obtaining picture pick-up device;To video file Pumping frame sampling is carried out, one group of video frame images is obtained;According to the pixel value of the pixel in one group of video frame images in one group of view Multiple target video frame images are determined in frequency frame image, wherein each target video frame image is used to indicate in monitoring area The middle object that there is movement;Target object detection is carried out to each target video frame image, obtains each target video frame image Characteristics of image, wherein characteristics of image be used for indicates exist move object in, the similarity between target object is greater than Object region where the object of first threshold;Determine that movement is special according to the characteristics of image of each target video frame image Sign, wherein motion feature is used to indicate the movement velocity that there is the object of movement in multiple target video frame images and movement side To;According to the characteristics of image of motion feature and each target video frame image, determine whether go out in multiple target video frame images Existing target object.That is, the video file to monitoring area carries out pumping frame sampling, one group of video frame images, root are obtained It determines to be used to indicate in monitoring area in one group of video frame images according to the pixel value of the pixel in one group of video frame images Multiple target video frame images of the middle object that there is movement, are determined further according to the characteristics of image of each target video frame image Motion feature, and then according to the characteristics of image of motion feature and each target video frame image, reach and automatically determines multiple targets The purpose that target object whether is had in video frame images not only greatly reduces the human cost of determining target object, and And the accuracy rate of determining target object is improved, solve the problems, such as the low efficiency being determined to target object, and then reach Propose the effect of high rodent infestation accuracy in detection.

As an alternative embodiment, in step s3, according to the pixel of the pixel in one group of video frame images Value determines that multiple target video frame images may include: every in one group of video frame images of acquisition in one group of video frame images The average pixel value of a pixel;Obtain the pixel of each pixel in each video frame images in one group of video frame images It is worth and the difference between corresponding average pixel value;Difference in one group of video frame images is met to the video frame images of predetermined condition It is determined as target video frame image.

Multiple mesh are being determined in one group of video frame images according to the pixel value of the pixel in one group of video frame images When marking video frame images, the pixel value of each pixel in available one group of video frame images, according to each pixel Calculated for pixel values goes out average pixel value, then obtain the pixel value of each pixel in one group of video frame images with it is corresponding average Difference between pixel value.

Optionally, the pixel of each pixel in each video frame images in one group of video frame images can also be obtained Value and the difference between background or the former frame of each video frame images.

After obtaining above-mentioned difference, judges whether difference meets predetermined condition, difference in one group of video frame images is expired The video frame images of sufficient predetermined condition are determined as target video frame image, to obtain multiple targets in one group of video frame images Video frame images.

As an alternative embodiment, obtaining each picture in each video frame images in one group of video frame images Difference between the pixel value of vegetarian refreshments and corresponding average pixel value may include: for each view in one group of video frame images Each pixel in frequency frame image executes following operation, wherein when executing following operation, each video frame images are considered as Each pixel is considered as current pixel point by current video frame image: D (x, y)=| f (x, y)-b (x, y) |, wherein (x, It y) is coordinate of the current pixel point in current video frame image, f (x, y) indicates the pixel value of current pixel point, b (x, y) table Show that the average pixel value of current pixel point, D (x, y) indicate between the pixel value and corresponding average pixel value of current pixel point Difference.

Obtain one group of video frame images in each video frame images in each pixel pixel value with it is corresponding When difference between average pixel value, each video frame images are considered as current video frame image, and each pixel is considered as Current pixel point can indicate coordinate of the current pixel point in current video frame image, for example, for current by (x, y) The video frame images upper left corner is origin, and wide direction is X-axis, and high direction is the coordinate of pixel in the coordinate system of Y-axis foundation, is passed through F (x, y) indicates the pixel value of current pixel point, and the average pixel value of current pixel point is indicated by b (x, y), is passed through D (x, y) Indicate current pixel point pixel value and corresponding average pixel value between difference, according to formula D (x, y)=| f (x, y)-b (x, y) | the difference between the pixel value of current pixel point and corresponding average pixel value is calculated, to reach by the above method To pixel value and the corresponding mean pixel of each pixel in each video frame images obtained in one group of video frame images The purpose of difference between value.

As an alternative embodiment, difference in one group of video frame images to be met to the video frame images of predetermined condition Being determined as target video frame image includes: that each pixel in each video frame images in one group of video frame images is held The following operation of row, wherein each video frame images are considered as current video frame image when executing following operation, by each pixel Point is considered as current pixel point:

Wherein, D (x, y) is expressed as the difference between the pixel value of current pixel point and corresponding average pixel value, T One preset threshold;Wherein, it is more than second that predetermined condition, which includes: the number of the pixel of M (x, y)=1 in target video frame image, Preset threshold.

In this embodiment, the video frame images that difference meets predetermined condition in one group of video frame images are being determined as mesh When marking video frame images, each video frame images are considered as current video frame image, and each pixel is considered as current pixel Point indicates that current video frame image, D (x, y) indicate pixel value and the corresponding mean pixel of current pixel point by M (x, y) Difference between value indicates the first preset threshold by T, if the number of the pixel of M (x, y)=1 is super in current video frame The second preset threshold is crossed, then current video frame image is determined as target video frame image, that is, then in current video frame image There are the objects of movement, are target video frame image, and otherwise, there is no the objects of movement in current video frame image.

Multiple target video frame images constitute movement destination image in above-mentioned one group of video frame images, can pass through form Merging pixel is calculated in student movement can obtain the object of all movements, as output result.

Optionally, the present embodiment in target video frame image exist movement object be detected as it is neural network based One group of video frame images can be sent into trained network model in advance, obtain all objects that there is movement by target detection With its confidence level, output of the characteristics of image as the network module of some confidence threshold value will be greater than.The network model used It may include but be not limited to single multi-target detection device (Single Shot MultiBox Detector, referred to as SSD), fast Fast region convolutional network (Faster Region-CNN, referred to as Faster-RCNN), feature pyramid network (Feature Pyramid Network, referred to as FPN) etc., no limitations are hereby intended.

As an alternative embodiment, in step s 5, the characteristics of image according to each target video frame image is true Making motion feature may include: to obtain object region pair represented by the characteristics of image with each target video frame image The target vector answered obtains multiple target vectors, wherein each target vector is for indicating corresponding target video frame figure There is movement velocity and the direction of motion of the object of movement when by object region as in;By multiple target vectors according to Time sequencing of each target video frame image in video file forms first object vector, wherein motion feature includes the One object vector;Or it obtains and object region corresponding two represented by the characteristics of image of each target video frame image Light stream figure is tieed up, multiple two-dimentional light stream figures are obtained, wherein each two dimension light stream figure includes in corresponding target video frame image There are movement velocity and the direction of motion of the object of movement when by object region;By multiple two-dimentional light stream figures according to every Time sequencing of a target video frame image in video file forms three-dimensional second object vector, wherein motion feature includes Three-dimensional second object vector.

Optionally, the characteristics of image of each target video frame image can calculate each for indicating object region The light stream (Optical flow or optic flow) of object region, obtains two dimension corresponding with the object region Light stream figure, and then obtain and the one-to-one multiple two-dimentional light stream figures of multiple target video frame images, wherein light stream is for describing The movement of observed object caused by movement relative to observer, surface or edge.Each of embodiment two dimension light stream figure Including exist in a corresponding target video frame image movement velocity of the object of movement when by object region and The direction of motion, that is, exist in target video frame image movement velocity of the object of movement when by object region and The direction of motion can be indicated by two-dimentional light stream figure.After obtaining multiple two-dimentional light stream figures, multiple two-dimentional light stream figures are pressed Three-dimensional second object vector is formed according to time sequencing of each target video frame image in video file, wherein each target Time sequencing of the video frame images in video file can indicate by time shaft, can by multiple two-dimentional light stream figures along when Between axis do and splice, obtain the second object vector, which is three-dimensional vector, using the three-dimensional vector as motion feature It is exported.

In the present embodiment, by for indicating that the object that there is movement in corresponding target video frame image is passing through The target vector of movement velocity and the direction of motion when crossing object region, or the image with each target video frame image Object region represented by feature corresponding two-dimentional light stream figure determines motion feature, which can be one-dimensional Vector is three-dimensional vector, determines motion feature according to the characteristics of image of each target video frame image to realize Purpose, and then according to the characteristics of image of motion feature and each target video frame image, it determines in multiple target video frame images Whether target object is had, achievees the purpose that automatically determine in multiple target video frame images whether have target object, Improve the accuracy rate of determining target object.

As a kind of optional example, by having merged, there are the detections of Moving Objects (target detection) and movement to above-mentioned The network of feature extraction exports characteristic pattern, and this feature figure has merged four dimensional vectors including vision and motion feature, wherein this four Dimensional vector can include but is not limited to time dimension, channel dimension, long dimension, high-dimensional.

As an alternative embodiment, in step s 6, according to motion feature and each target video frame image Characteristics of image determines that it may include: by motion feature and each that target object whether is had in multiple target video frame images The characteristics of image of target video frame image is input in preparatory trained neural network model, obtain Object identifying as a result, its In, Object identifying result is for indicating whether have target object in multiple target video frame images.

In this embodiment, in the characteristics of image according to motion feature and each target video frame image, multiple mesh are determined It, can be by the image of motion feature and each target video frame image spy when whether having target object in mark video frame images Sign is input in preparatory trained neural network model, obtains Object identifying as a result, the neural network model namely classification net Network model, according to there may be the characteristics of image sample of the target object of movement, motion feature sample and being used to indicate target The data of object are trained initial neural network model, and for determining in video frame images whether have target object Model.Object identifying result namely classification results differentiate as a result, for indicating whether occur in multiple target video frame images There is target object.

As an alternative embodiment, the characteristics of image of motion feature and each target video frame image is input to In preparatory trained neural network model, obtaining Object identifying result may include: to pass through each characteristics of image including volume The neural net layer structure of lamination, regularization layer and activation primitive layer, obtains multiple first eigenvectors;By multiple fisrt feature Vector is merged with motion feature, obtains second feature vector;Second feature vector is input to full articulamentum to classify, Obtain the first classification results, wherein neural network model includes neural net layer structure and full articulamentum, Object identifying result packet The first classification results are included, the first classification results are for indicating whether have target object in multiple target video frame images;Or Each characteristics of image is passed through the first nerves network layer structure including convolutional layer, regularization layer and activation primitive layer by person, is obtained Multiple first eigenvectors;Motion feature is passed through to the nervus opticus network layer including convolutional layer, regularization layer, activation primitive layer Structure obtains second feature vector;Multiple first eigenvectors are merged with second feature vector, obtain third feature to Amount;Third feature vector is input to full articulamentum to classify, obtains the second classification results, wherein neural network model packet First nerves network layer structure, nervus opticus network layer structure and full articulamentum are included, Object identifying result includes the second classification knot Fruit, the second classification results are for indicating whether have target object in multiple target video frame images.

Optionally, the overall structure of neural network model can be divided into convolutional layer, regularization layer, activation primitive layer, Quan Lian Connect layer, wherein convolutional layer is made of several convolution units, and the parameter of each convolution unit is best by back-propagation algorithm What change obtained;Regularization layer can be used for preventing the over-fitting of neural network model training, and activation primitive layer can will be non-linear Network is introduced, full articulamentum plays the role of classifier in entire convolutional neural networks.

In the present embodiment, the characteristics of image of motion feature and each target video frame image is being input to preparatory training In good neural network model, when obtaining Object identifying result, each characteristics of image can be passed through includes convolutional layer, regularization Layer and activation primitive layer neural net layer structure, obtain multiple first eigenvectors, by multiple first eigenvector with it is upper It states motion feature to be merged, to obtain second feature vector, wherein motion feature is motion in one dimension feature.

As a kind of optional amalgamation mode, multiple first eigenvectors and motion feature can be spliced (or For combination), obtain second feature vector.

After obtaining second feature vector, second feature vector is input to full articulamentum and is classified, that is, logical complete Articulamentum classifies to second feature vector, to obtain the first classification results, wherein the neural network model of the embodiment Including above-mentioned neural net layer structure and above-mentioned full articulamentum, the first classification results are for indicating in multiple target video frame images Whether the Object identifying of target object is had as a result, for example, whether to have mouse in multiple target video frame images Classification results.

Optionally, above-mentioned that each characteristics of image is passed through to the nerve net including convolutional layer, regularization layer and activation primitive layer Network layers structure obtains multiple first eigenvectors, and multiple first eigenvectors are merged with motion feature, obtains the second spy Vector is levied, second feature vector is input to full articulamentum and is classified, the method for obtaining the first classification results can obtain Target vector corresponding with object region represented by the characteristics of image of each target video frame image, obtains multiple targets Vector, by multiple target vectors according to time sequencing of each target video frame image in video file form first object to It is executed after amount.

Optionally, the characteristics of image of motion feature and each target video frame image is being input to preparatory trained mind When through in network model, obtaining Object identifying result, each characteristics of image is passed through including convolutional layer, regularization layer and activation letter Several layers of first nerves network layer structure, obtains multiple first eigenvectors;Above-mentioned motion feature is passed through including convolutional layer, just The nervus opticus network layer structure for then changing layer, activation primitive layer, obtains second feature vector.It is obtaining first eigenvector and is obtaining To after second feature vector, multiple first eigenvectors are merged with second feature vector, obtain third feature vector.

As another optional amalgamation mode, multiple first eigenvectors and second feature vector can be spliced (or being combination), obtains third feature vector.

After obtaining third feature vector, third feature vector is input to full articulamentum and is classified, to obtain Second classification results, wherein the neural network model of the embodiment includes first nerves network layer structure, nervus opticus network layer Structure and full articulamentum, Object identifying result include the second classification results, and second classification results are for indicating multiple target views Whether target object is had in frequency frame image, for example, whether to have the classification of mouse in multiple target video frame images As a result.

Optionally, above-mentioned that each characteristics of image is passed through to the first mind including convolutional layer, regularization layer and activation primitive layer Through network layer structure, multiple first eigenvectors are obtained, it includes convolutional layer, regularization layer, activation primitive that motion feature, which is passed through, The nervus opticus network layer structure of layer, obtains second feature vector, and multiple first eigenvectors and second feature vector are carried out Fusion, obtains third feature vector, third feature vector is input to full articulamentum and is classified, the second classification results are obtained Method can obtain two-dimentional light corresponding with object region represented by the characteristics of image of each target video frame image Flow graph obtains multiple two-dimentional light stream figures, by multiple two-dimentional light stream figures according to each target video frame image in video file Time sequencing executes after forming three-dimensional second object vector.

As another optional example, the characteristics of image of motion feature and each target video frame image is input to pre- First in trained neural network model, obtaining Object identifying result includes: that each characteristics of image is successively passed through multiple pieces, is obtained To multiple first eigenvectors, wherein convolution operation, the canonical successively executed on convolutional layer can be inputted to block in each piece Change the regularization operation on layer, the operation of the activation on activation primitive layer;Multiple first eigenvectors are spelled with motion feature It connects, obtains second feature vector;Second feature vector is input to full articulamentum, exports to obtain the first classification by full articulamentum As a result, wherein neural network model includes multiple pieces and full articulamentum, and Object identifying result includes the first classification results, and first Classification results are for indicating whether have target object in multiple target video frame images;Or successively by each characteristics of image By multiple first pieces, multiple first eigenvectors are obtained, wherein can successively hold to first piece of input in each first piece The regularization operation in convolution operation, regularization layer on row convolutional layer, the operation of the activation on activation primitive layer;By motion feature Successively pass through multiple second pieces, obtain second feature vector, wherein second piece of input can successively be held in each second piece The regularization operation in convolution operation, regularization layer on row convolutional layer, the operation of the activation on activation primitive layer;By multiple first Feature vector is spliced with second feature vector, obtains third feature vector;Third feature vector is input to full articulamentum, Export to obtain the second classification results by full articulamentum, wherein neural network model includes multiple first pieces, it is multiple second piece and Full articulamentum, Object identifying result include the second classification results, and the second classification results are for indicating multiple target video frame images In whether have target object.

In the present embodiment, each characteristics of image can also be handled by block.Can by each characteristics of image according to It is secondary to pass through multiple pieces, multiple first eigenvectors are obtained, the input of block can successively be executed on convolutional layer in each piece Convolution operation is operated in the regularization operation on regularization layer and the activation on activation primitive layer.Obtaining multiple first After feature vector, multiple first eigenvectors are spliced with motion feature, to obtain second feature vector.It is obtaining After second feature vector, second feature vector is input to full articulamentum and is classified, exports to obtain by full articulamentum One classification results, wherein the neural network model of the embodiment includes multiple pieces and full articulamentum, and Object identifying result includes the One classification results, first classification results are used for indicating whether have target object in multiple target video frame images, for example, For the classification results for whether having mouse in multiple target video frame images.

Optionally, which is handled each characteristics of image by first piece, and each characteristics of image is successively passed through Multiple first pieces are crossed, multiple first eigenvectors are obtained, first piece of input can successively be executed and rolled up in each first piece Convolution operation on lamination, in the regularization operation on regularization layer and the activation operation on activation primitive layer.The implementation Example can also be handled motion feature by second piece, and motion feature is successively passed through to multiple second pieces, obtain the second spy Vector is levied, can successively execute convolution operation on convolutional layer to second piece of input in each second piece, in regularization layer On regularization operation and on activation primitive layer activation operation.Obtain multiple first eigenvectors and second feature to After amount, multiple first eigenvectors and second feature vector are spliced, obtain third feature vector, finally by third spy Sign vector is input to full articulamentum and classifies, and exports to obtain the second classification results by full articulamentum, wherein the embodiment Neural network model includes multiple first pieces, multiple second piece and full articulamentums, and Object identifying result includes the second classification results, Second classification results are for indicating whether have target object in multiple target video frame images, for example, being multiple targets Whether the classification results of mouse are had in video frame images.

As an alternative embodiment, carry out pumping frame sampling to video file, obtaining one group of video frame images can be with Include: that equally spaced pumping frame sampling is carried out to the video sequence in video file, obtains one group of video frame images.

Video file includes video sequence, can carry out pumping frame sampling to video file, obtain one group of video frame images When, equally spaced pumping frame sampling is carried out to the video sequence in video file, obtains one group of video frame images, to reduce to mesh The operand for the algorithm that mark object is determined, and then target object whether is quickly had in multiple target video frames, it improves Efficiency that target object is determined.

As an alternative embodiment, obtaining picture pick-up device can wrap the video file that monitoring area is shot It includes: obtaining the video file that infrared low-light night vision camera shoots monitoring area, wherein the video frame in video file Image is the image taken by infrared low-light night vision camera.

Picture pick-up device can be camera, for example, being infrared low-light night vision camera, the infrared low-light night vision camera band There is infrared illumination function.Target area is shot by infrared low-light night vision camera, obtains video file, video text Video frame images in part are the image taken by infrared low-light night vision camera.

Optionally, picture pick-up device further includes but is not limited to: mobile detection function, network savvy (such as wifi networking) and clear Spend (such as larger than 1080p) configuration.

As an alternative embodiment, whether had in determining multiple target video frame images target object it Afterwards, in the case where target object being had in determining multiple target video frame images, determine target object multiple Position in target video frame image;Position is shown in multiple target video frame images.

After whether having target object in determining multiple target video frame images, multiple target videos are being determined In the case where having target object in frame image, it may further determine that target object in multiple target video frame images Position for example, determining position of the mouse in multiple target video frame images, and then position is shown in multiple target video frames In image, for example, showing the information such as the icon for being used to indicate position, text in multiple target video frame images.

Optionally, the information such as time, the zone of action in monitoring area that can also obtain target object appearance, by mesh Mark the position of object, the time, the specific zone of action in monitoring area, in motion frequency, motion track of monitoring area etc. Information is exported to front end, the front end namely display unit, and the information such as time, zone of action that target object occurs can shown It is shown in interface, determines that target object causes to be to the inefficiency that target object is determined so as to avoid artificial Problem.

Optionally, in the case where having target object in determining multiple target video frame images, report can be sent Alert information is to front end, which, which is used to indicate in monitoring area, has target object, so that related control personnel takes Control measure, to improve the efficiency prevented and treated target object.

As an alternative embodiment, the determination of target object can be executed by setting local server, without connecting Cloud Server is connect, inside can be realized above-mentioned operation and visualization, avoid operation end on Cloud Server, have calculating money Problem on source, in transmission, the problem for causing entire frame efficiency more low, is determined target object to improve Efficiency.

The determination method of above-mentioned target object is intended to the technology of application image identification, blending image feature and motion feature, Whether there is target object in automatic detection monitor video, target object is positioned and is tracked, the shifting of target object can be generated Dynamic rail mark and motion frequency in each monitoring area, whole process is all algorithm realization, without additional human cost;In addition, The embodiment is not necessarily to determine the target object in monitoring area by drop target capture device, carries out without cost manpower Observation, not only greatly reduces the human cost of monitoring objective object, improves the efficiency being determined to target object, in turn Facilitate the work further prevented and treated target object.

It is illustrated below with reference to determination method of the optional embodiment to above-mentioned target object.Specifically with target pair As being illustrated for mouse.

In this alternative embodiment, the determination method of target object be may comprise steps of:

Step S1 obtains the video file that infrared low-light night vision camera takes.

Step S2 judges in video file with the presence or absence of moving object.

Step S3, if there is moving object, then there are the video clips of moving object for extraction.

Step S4, to there are the video clips of moving object to carry out characteristics of image and behavioral characteristics extraction.

Step S5 judges whether moving object is mouse according to the characteristics of image and behavioral characteristics that extract.

Step S6, if it is judged that be it is yes, then issue prompt information.

In the present embodiment, the video file taken using infrared low-light night vision camera is obtained;Judge video file In whether there is moving object;If there is moving object, then there are the video clips of moving object for extraction;To there are moving objects The video clip of body carries out characteristics of image and behavioral characteristics extract;According to characteristics of image and behavioral characteristics the judgement movement extracted Whether object is mouse;If it is judged that be it is yes, then issue prompt information, target object be determined to solve The problem of low efficiency, and then achieved the effect that propose high rodent infestation accuracy in detection.

The technical solution of the present embodiment can be used as a kind of fusion visual signature and the mouse of track characteristic suffers from video surveillance side Method can be applied in several scenes for detecting in the video taken with the presence or absence of mouse, be taken the photograph by infrared low-light night vision As the video file of head shooting current environment, moving object is then judged whether there is, if there is moving object, then by mentioning It takes the video clip of moving object to carry out feature identification, further judges to extract whether moving object is mouse, if it is judged that It is mouse, then issues prompt information, prompt information can be shows text on the screen, is also possible to make a sound prompt letter Breath, is also possible to a plurality of types of prompt informations such as bright light or flashing.

It should be noted that monitoring camera is taken the photograph using outer lll night vision in the technical solution of the embodiment of the present invention As head, in addition, the treatment processes such as its judgement, extraction are carried out in local server, without transmitting data to long-range clothes Device be engaged in handle, it is possible to reduce volume of transmitted data improves monitoring efficiency.

Optionally, after issuing prompt information, position of the moving object in video file in every frame picture is determined;It will Preset mark is superimposed upon at the corresponding position of every frame picture and is shown in front-end interface.

After sending has the prompt of mouse, determines position of the mouse in video file in every frame picture, then will preset Label be superimposed upon at the corresponding position of every frame picture and show, preset mark can be green or red rectangle frame, often The position of mouse is marked with rectangle frame in frame picture, to facilitate user that can view the position of mouse in time and often haunt Region.

Optionally, judge that whether there is moving object in video file includes: to carry out to the video sequence in video file Equally spaced pumping frame sampling, obtains sampled video frame;It is examined by Detection dynamic target algorithm or target neural network based Method of determining and calculating judges whether there is moving object in sampled video frame image.

When whether there is moving object in judging video file, equally spaced pumping frame can be carried out to video sequence and adopted Then sample judges whether there is moving object in sampled video frame to reduce the operand of algorithm, dynamic mesh can be used when judging Any one in mark detection algorithm or algorithm of target detection neural network based in some cases can also the two It is used in mixed way.

Optionally, judge that it includes: logical for whether having moving object in sampled video frame image by Detection dynamic target algorithm Cross Dk(x, y)=| fk(x,y)-bk(x, y) | calculate the difference of present frame and background or former frame;Pass throughJudge whether there is moving object, wherein it is former that (x, y), which is with the image upper left corner, Point, wide direction are X-axis, and high direction is the coordinate of pixel in the coordinate system of Y-axis foundation, and k is the index of present frame, and f expression is worked as Previous frame, b indicate that background or previous frame, M (x, y) are moving image, and T is threshold value.

If M (x, y) indicates moving target for 1, the pixel of all X (x, y) constitutes movement destination image, by form Merging pixel is calculated in student movement can obtain the target of all movements.

Optionally, judge that moving object whether be mouse includes: that will mention according to the characteristics of image and behavioral characteristics that extract The characteristics of image and behavioral characteristics got are input in preparatory trained neural network model, are carried out Model checking, are obtained mould Type exports result;Result, which is exported, according to model judges whether moving object is mouse.

Model can be carried out to the characteristics of image and behavioral characteristics extracted by preparatory trained neural network model Differentiate, model is obtained previously according to a large amount of sample training, and whether a large amount of sample includes having in picture and the picture always The label of mouse can also include the label of the mouse quantity in the picture, model can be made more smart so in some cases Really.

The technical solution of the present embodiment, which can be applied, needs to monitor whether the application scenarios of damaged by rats in kitchen, dining room etc. In, it also can be used in the place that the indoor and outdoors such as hotel industry school, laboratory, hospital require environmental sanitation, in mouse In evil preventing and controlling, mouse detection and tracking is carried out using the image recognition technology of this alternative embodiment, uses independent one Device, without placing mousetrap mouse cage, is seen by monitoring camera in the monitoring for locally completing mouse trouble without cost manpower It surveys, will monitor the plague of rats becomes the process work of high efficient full automatic, not only greatly reduces the human cost of the monitoring plague of rats, while quasi- True rate is high, the convenient supervision to plague of rats health, and provides trace information, facilitates further mouse killing working.

The technical solution of the embodiment of the present invention additionally provides a kind of preferred embodiment, below with reference to the preferred embodiment The technical solution of the embodiment of the present invention is illustrated.

The preferred embodiment of the present invention is intended to the technology of application image identification, merges vision and image sequence characteristic, automatic to examine Survey monitor video in whether have mouse, mouse is positioned and is tracked, and generate mouse motion profile route and each region Motion frequency, whole process be all algorithm realization, without additional human cost, and be an independent device, be not necessarily to Cloud Server is connected, all operation and visualization can be achieved in inside.

A kind of mouse is provided in a preferred embodiment of the invention suffers from video monitoring device.The device is totally divided into following Component: infrared low-light night vision camera (being connected with data processing module), data processing module are (with infrared low-light night vision camera It is connected with front end display unit) and front end display unit, principle is as follows when which works:

Infrared low-light night vision camera is responsible for acquiring scene video sequence, and data processing module receives video sequence and examines Whether there is or not mouse in survey video, if detecting mouse, the range of information such as the position of mouse are exported to front end display interface, front end Display interface shows position, time of occurrence, zone of action and the alarm that can carry out mouse trouble immediately of mouse.

Above-mentioned data processing module can be divided into video acquisition module, video processing module and memory module.Fig. 4 is basis The schematic diagram of a kind of each module data connection of the preferred embodiment of the present invention, as shown in figure 4, video acquisition module passes through ARM plate Video data is acquired, and is pre-processed by video pre-filtering module, video processing module reads in trained model and exists Video processing is carried out according to deep learning algorithm in embedded gpu processor, if deep learning network model detects some Fractional time has mouse, then stores the segment and corresponding testing result to memory module, memory module is a series of by this Information is exported to front end.

Fig. 5 is the schematic illustration that a kind of mouse according to the preferred embodiment of the invention suffers from detection method.As shown in figure 5, should The principle that mouse suffers from video detecting method may include following components: pretreatment, target detection, and motion feature extracts and classification Network.The input of this method is original video sequence.This method is specifically described below.

Pretreatment:

Pretreatment may include two steps: taking out frame and dynamic detection, carries out before this to original video sequence equally spaced Frame sampling is taken out, the operand of algorithm is reduced, then target detection is carried out using algorithm of target detection, judges whether there is fortune in image Animal body, without subsequent detection, if there is moving object, will there is the video clip of moving object if without motion object It is sent into subsequent module.During target detection, each frame of pretreated video sequence is detected, there may be The position acquisition characteristics of image (visual information in such as corresponding detection block in the position) of mouse, and mould is extracted by motion feature Information between each video image frame is carried out fusion and feature extraction, prevents the object detector of single frames from judging by accident by block The case where, sorter network then is inputted by the motion feature of extraction and with characteristics of image, discriminates whether it is mouse by sorter network, If mouse, then the hough transform frame by mouse in each frame position is transmitted to front end display interface.

Target detection:

In a preferred embodiment of the invention, according to two kinds of algorithms of specific machine computational resource allocation: dynamic object inspection It surveys and target detection neural network based, the former arithmetic speed is fast, the latter accuracy and robust low to machine configuration requirement Property.

1) Detection dynamic target algorithm includes background subtraction and frame difference method, using following formula (1), calculating present frame and background Or the difference of former frame:

Dk(x, y)=| fk(x,y)-bk(x,y)| (1)

In above formula, (x, y) is using the image upper left corner as origin, and wide direction is X-axis, and high direction is the coordinate system that Y-axis is established The coordinate of middle pixel, k are the index of present frame, and f represents present frame, and b represents background or previous frame.Sentenced using formula (2) It is disconnected to whether there is moving target:

Wherein, M (x, y) is moving image, and T is threshold value, if M (x, y) indicates moving target for 1, all X (x, y) Pixel constitutes movement destination image, the target of all movements can be obtained by morphology operations merging pixel, as the mould The output of block.

2) algorithm of target detection neural network based by picture input in advance trained network model, obtain it is all can Can target and its confidence level, greater than output of the detection block as the module of some confidence threshold value.The network model used Including but not limited to SSD, Faster-RCNN, FPN etc..Fig. 6 is a kind of Faster-RCNN network of the preferred embodiment of the present invention The schematic diagram of model.As shown in Figure 6, wherein conv is convolutional layer, is drawn in input by convolution kernel (being a matrix) Window draws window position and matrix according to formula (3) phase dot product to each input, and as a result F is defeated as the feature of this stroke of window position Out.

RPN is that region proposes network, can propose that a series of candidate frame, the pond ROI pooling layer mention convolutional layer Area maps of the characteristic pattern under the coordinate that RPN the is exported rectangle frame fixed at size (w, h), feeding is made of full articulamentum Classifier and frame return device, frame return output mouse possibility coordinate position, classifier output be the position mouse Confidence level.

Motion feature extracts:

Because the movement of object is detection block continuous, that motion feature extraction algorithm is first obtained according to each frame, calculate The correlation of detection block between frame and frame, the big detection block of correlation are considered same object, carry out to the detection block of each frame Matching, obtains the range of motion picture of object, and the feature of motion sequence is finally extracted using the feature extraction network of 3D.

Sorter network:

By the visual information and motion feature fusion in target detection frame, it is sent into the network model of designed classification, is used In the sequence of pictures for screening out non-mouse, false alarm rate is reduced, result is sent into front end display interface, shows the detection block and rail of mouse Mark.

In embodiments of the present invention, for whole frame, can with but be not limited by target detection and sorter network It is identified to achieve the purpose that detect, to save frame layout cost.

The embodiment of the present invention is proposed using image recognition algorithm, the mouse in automatic identification monitor video, without placing Mousetrap mouse cage, without spending manpower to be observed, will monitor the plague of rats becomes the process work of high efficient full automatic, not only subtracts significantly The human cost of the monitoring plague of rats is lacked, while accuracy rate is high, the convenient supervision to rear kitchen plague of rats health, at the same time it can also provide Personnel selection deratization tool placement location is convenient in the movable track of mouse, facilitates work of further removing the evil.

Through the above description of the embodiments, those skilled in the art can be understood that according to above-mentioned implementation The method of example can be realized by means of software and necessary general hardware platform, naturally it is also possible to by hardware, but it is very much In the case of the former be more preferably embodiment.Based on this understanding, technical solution of the present invention is substantially in other words to existing The part that technology contributes can be embodied in the form of software products, which is stored in a storage In medium (such as ROM/RAM, magnetic disk, CD), including some instructions are used so that a terminal device (can be mobile phone, calculate Machine, server or network equipment etc.) method that executes each embodiment of the present invention.

A kind of monitoring device of target object is additionally provided in the present embodiment, is applied to local server, which uses In realizing above-described embodiment and preferred embodiment, the descriptions that have already been made will not be repeated.As used below, term The combination of the software and/or hardware of predetermined function may be implemented in " module ".Although device is preferably described in following embodiment It is realized with software, but the realization of the combination of hardware or software and hardware is also that may and be contemplated.

Fig. 7 is the structural block diagram one of the monitoring device of target object according to an embodiment of the present invention, as shown in fig. 7, the dress It sets and includes:

Receiving module 72, the target image sent for receiving target monitoring equipment by local net network, wherein mesh Logo image is the image that target monitoring equipment is shot in the case where the monitored picture for detecting target monitoring equipment changes;

Identification module 74, for carrying out recongnition of objects to target image;

Determining module 76 in the case where for including target object in identifying target image, determines that target is known Other result, wherein target identification result is for indicating target object occur in the monitoring area of target monitoring equipment.

Optionally, receiving module 72 can be used for executing above-mentioned steps S202, and identification module 74 can be used for executing above-mentioned Step S204, determining module 76 can be used for executing above-mentioned steps S206.

By above-mentioned each module, local server receives the target figure that target monitoring equipment is sent by local net network Picture, wherein target image is target monitoring equipment in the case where the monitored picture for detecting target monitoring equipment changes The image of shooting;Local server carries out recongnition of objects to target image;It include target in identifying target image In the case where object, local server determines target identification result, wherein target identification result is for indicating in target monitoring There is target object in the monitoring area of equipment, the monitoring method for solving target object in the related technology, which can not be stablized, to be held The lower problem of capable and target object recognition efficiency, ensure that the stability that monitoring executes, reduces transmission time, mention High transmission speed, and then improve monitoring efficiency.

Optionally, above-mentioned apparatus further include: first obtains module, for obtaining after determining target identification result Target video comprising target image, wherein target video includes the video that target monitoring equipment is shot in the target time period Data, target time section are the period at the first moment to the second moment, and the first moment was that target monitoring equipment detects monitoring At the time of picture starts variation, at the time of the second moment was that target monitoring equipment detects that monitored picture terminates variation.

Optionally, receiving module 72 includes: receiving unit, and the first acquisition module includes: first acquisition unit, wherein

First receiving unit, for receiving target monitoring equipment by local net network in the target time period every mesh Mark the target image that time interval is sent;

First acquisition unit, for being more than not receive the mesh of target monitoring equipment transmission the time of target time interval In the case where logo image, the target video comprising target image is obtained.

Optionally, the first acquisition module includes:

Transmission unit, for sending request message to target monitoring equipment, wherein request message includes target for requesting The video of image;Second receiving unit, for receiving the target video of target monitoring equipment response request message transmission;Alternatively,

Unit is transferred, the video data concentration for saving from local server transfers target video, wherein video data The video for the target monitoring equipment shooting that collection is uploaded comprising target monitoring equipment.

Optionally, above-mentioned apparatus further include: sending module, for obtain comprising target image target video after, It sends a warning message to output end, wherein warning information is used for there is target object in the monitoring area of target monitoring equipment It is alerted, warning information includes at least one of: target image, target video.

Optionally, above-mentioned apparatus further include: the first determining module, for being determined and target video pair according to target video The target information answered, wherein target information includes at least one of: travel path of the target object in monitoring area, mesh Mark object enters the invasion point of monitoring area, and target object concealing a little in monitoring area, target object is in monitoring area Activity time, the characteristics of objects of target object, warning information further include: target information.

Optionally, above-mentioned apparatus further include: mark module, for marking target object and and mesh in target video Mark the corresponding target information of video, wherein the case where target information includes travel path of the target object in monitoring area Under, travel path is by a string on motion track in monitoring area of target object there are the punctuates of number to indicate, mark Number on point is sequentially increased along the moving direction of target object.

Optionally, above-mentioned apparatus further include: generation module, for including target object in monitoring area in target information Travel path in the case where, according to travel path generate prompt information, wherein prompt information for prompt eliminate target object Mode, warning information further includes prompt information.

Optionally, above-mentioned apparatus further include:

Second obtains module, for before sending a warning message to output end, obtaining the target knot of multiple monitor videos Fruit, wherein multiple monitor videos are the monitor video that multiple monitoring devices in target area upload, and multiple monitoring devices include Target monitoring equipment, objective result are target information corresponding with multiple monitor videos;

Second determining module determines one or more positions in target area for the objective result according to acquisition Information, wherein one or more location informations include at least one of: target object enters the invasion point of target area, mesh Mark object concealing a little in target area, target object in target area frequency of occurrence be more than targeted number threshold value and/or Time of occurrence is more than the location point of object time threshold value.

Optionally, identification module 74 includes:

Optionally, above-mentioned apparatus further include: removing module is used for after carrying out recongnition of objects to target image, In the case where not including target object in identifying target image, the target video of delete target monitoring device upload, wherein Target video includes the video data that shoots in the target time period of target monitoring equipment, and target time section was the first moment to the The period at two moment, at the time of the first moment was that target monitoring equipment detects that monitored picture starts variation, the second moment was At the time of target monitoring equipment detects that monitored picture terminates variation.

The monitoring device of another target object is additionally provided in the present embodiment, is applied to monitoring device, which uses In realizing above-described embodiment and preferred embodiment, the descriptions that have already been made will not be repeated.As used below, term The combination of the software and/or hardware of predetermined function may be implemented in " module ".Although device is preferably described in following embodiment It is realized with software, but the realization of the combination of hardware or software and hardware is also that may and be contemplated.

Fig. 8 is the structural block diagram two of the monitoring device of target object according to an embodiment of the present invention, as shown in figure 8, the dress It sets and includes:

Detection module 82 carries out mobile detection for the monitoring area to monitoring device, wherein mobile detection is for detecting Whether the monitored picture of monitoring device changes;

Module 84 is obtained, in the case where detecting that monitored picture changes, monitoring area to be clapped in acquisition The target image taken the photograph;

Uploading module 86, for target image to be uploaded to local server by local net network, wherein target image It is used to indicate local server and recongnition of objects is carried out to target image.

Optionally, detection module 82 can be used for executing above-mentioned steps S302, and acquisition module 84 can be used for executing above-mentioned Step S304, uploading module 86 can be used for executing above-mentioned steps S306.

By above-mentioned module, monitoring device carries out mobile detection to the monitoring area of monitoring device, wherein mobile detection is used Whether change in the monitored picture of detection monitoring device;In the case where detecting that monitored picture changes, monitoring is set It is standby to obtain the target image shot to monitoring area;Target image is uploaded to by monitoring device by local net network Local server, wherein target image is used to indicate local server and carries out recongnition of objects to target image, solves phase The monitoring method of target object in the technology of pass can not stablize the lower problem of recognition efficiency of execution and target object, protect The stability that monitoring executes has been demonstrate,proved, has reduced transmission time, improves transmission speed, and then improve monitoring efficiency.

Optionally, uploading module 86 includes: the first uploading unit, for shooting monitoring device every target time interval To target image be uploaded to local server, wherein target time section be the first moment to the second moment period, first At the time of moment is that monitoring device detects that monitored picture starts variation, the second moment was that monitoring device detects monitored picture knot At the time of Shu Bianhua.

Optionally, uploading module 86, be also used to by target image by local net network be uploaded to local server it Afterwards, in the case where detecting that monitored picture terminates variation, target video is uploaded to local server, wherein target video Comprising the video data that monitoring device is shot in the target time period, target time section is the time at the first moment to the second moment Section, at the time of the first moment was that monitoring device detects that monitored picture starts variation, the second moment was that monitoring device detects prison At the time of control picture terminates variation.

Optionally, above-mentioned apparatus further include:

First determining module, for determining first before target video is uploaded to local server by monitoring device Moment and the second moment;

First read module, for reading out target video from the memory of monitoring device, wherein target video is monitoring The video data at third moment to the 4th moment of equipment shooting, the third moment be the first moment before, be separated by with the first moment At the time of first time period, the 4th moment is after the second moment, at the time of being separated by the second time interval with the second moment.

Optionally, above-mentioned apparatus further include:

Second read module, for detecting monitored picture knot before target video is uploaded to local server In the case where Shu Bianhua, monitoring device reads out target video from the memory of monitoring device;

Writing module, for target video to be written in the storage unit of monitoring device.

Optionally, uploading module 86 includes:

First detection unit, for during target video is uploaded to local server, detecting local area network net Network is in abnormality;

Storage unit, for saving the upload progress of target video;

Second uploading unit, in the case where detecting that local net network is in normal condition, according to the upper of preservation Degree of coming into continues to upload target video to local server.

Optionally, detection module 82 includes:

Second detection unit carries out mobile inspection for monitoring area of the camera by monitoring device to monitoring device It surveys.

A kind of monitoring system of target object is additionally provided in the present embodiment, and Fig. 9 is mesh according to an embodiment of the present invention The structural block diagram of the monitoring system of object is marked, as shown in figure 9, the system includes: target monitoring equipment 92 and local server 94, Target monitoring equipment 92 is connected with local server 94 by local net network, and above-mentioned target monitoring equipment 92 may include above-mentioned The monitoring device of the target object applied to monitoring device of any one, above-mentioned local server 94 may include any of the above-described The target object applied to local server monitoring device.

Optionally, target monitoring equipment 92 carries out mobile detection for the monitoring area to target monitoring equipment, wherein Mobile detection is for detecting whether the monitored picture of target monitoring equipment changes;Detecting that monitored picture is changed In the case of, obtain the target image shot to monitoring area;Target image is uploaded to this by local net network Ground server;

Local server 94, the target image sent for receiving target monitoring equipment by local net network;To mesh Logo image carries out recongnition of objects;In the case where including target object in identifying target image, determine that target is known Other result, wherein target identification result is for indicating occur target object in monitoring area.

Optionally, target monitoring equipment 92 is also used to:

Target monitoring equipment is uploaded to this every the target image that target time interval takes in the target time period Ground server, wherein target time section is the period at the first moment to the second moment, and the first moment detected for monitoring device At the time of monitored picture starts variation, at the time of the second moment was that monitoring device detects that monitored picture terminates variation.

Optionally, target monitoring equipment 92 is also used to:

In the case where detecting that monitored picture terminates variation, target video is uploaded to local server, wherein target Video includes the video data that monitoring device is shot in the target time period, and target time section is the first moment to the second moment Period, at the time of the first moment was that target monitoring equipment detects that monitored picture starts variation, the second moment was target monitoring At the time of equipment detects that monitored picture terminates variation.

Optionally, above system further includes output end, and output end is connected with local server 94, wherein

Local server 94 is also used to receive the target video of target monitoring equipment upload;It is determined according to target video Target information corresponding with target video, wherein target information includes at least one of: target object is in monitoring area Travel path, the invasion point into monitoring area of target object, target object concealing a little in monitoring area, target object Activity time in monitoring area, the characteristics of objects of target object;It include target object in monitoring area in target information Travel path in the case where, according to travel path generate prompt information, wherein prompt information for prompt eliminate target object Mode;It sends a warning message to output end, wherein warning information is used for there is mesh in the monitoring area of target monitoring equipment Mark object is alerted, and warning information includes at least one of: target image, target video, target information, prompt information;

Output end, for receiving warning information;At least one of is shown on the screen of output end: target image, mesh Mark video, target information, prompt information.

Optionally, local server 94 is also used to:

In the case where not including target object in identifying target image, the target that delete target monitoring device uploads is regarded Frequently, wherein target video includes the video data that target monitoring equipment is shot in the target time period, target time section first The period at moment to the second moment, at the time of the first moment was that target monitoring equipment detects that monitored picture starts variation, the At the time of two moment were that target monitoring equipment detects that monitored picture terminates variation.

It is described in detail below with reference to alternative embodiment of the present invention.

Alternative embodiment of the present invention provides a kind of monitoring framework of target object, and Figure 10 is optional implementation according to the present invention The schematic diagram of the monitoring framework of the target object of example proposes a kind of system architecture as shown in Figure 10, monitor internal and external environment and Harmful organism movement information.The system is without accessing internet, it is only necessary to local net network, while also have can rapid deployment spy Sign needs to be used to acquire data for data back and calculating and video monitoring equipment in curstomer's site deployment services device, with And LAN environment is transmitted for data.All subsequent calculating analyses are completed in local server.This system has also combined worm Plague of rats monitoring and worm rat plague control, form benign closed loop, play for the work of actual worm rat plague control and globally assist to make With.

The system architecture includes following hardware:

Camera monitors the variation of picture for using the mode of mobile detection;When having any variation generation, that is, record Video, interception picture, are sent to local server.

Local server, for receiving pictures and video data from camera shooting head end, image is completed in data storage Identification, judges whether it is the path analysis of haunting of the worm plague of rats and mouse, cockroach.

Output end is directly connected to server, for showing the monitoring result of harmful organism at the scene.Has network condition When, it also may be selected in APP or other information push mode, to user's display data.Output end can be a display terminal.

The monitoring method of the target object for the monitoring framework for being applied to above-mentioned target object is illustrated below.

The monitoring method of target object may comprise steps of:

(1) data collection steps (acquisition video and pictures):

In places such as rear kitchens, the preferable position in the suitable visual field is selected, disposes video monitoring equipment, kitchen key is set after acquisition The video data applied, to observe insects, muroid haunts situation.One visual actual conditions of indoor environment, deployment multiple groups monitoring Equipment.In view of mouse is haunt the characteristics of at night, video monitoring equipment need to have infrared night viewing function.

Video monitoring equipment uses the mode of mobile detection, (the ratio when any variation occurs for the image content produced When being flown into if any mouse appearance, cockroach appearance or foreign matter), it (generally can be to video preprocessor by the video write-in SD card in the period Record and be delayed 5 seconds, video is enabled to record complete one section of movement), video data is uploaded to local server in time.

Camera possesses broken string and resumes function, when wireless network environment is unstable, can also guarantee that video is complete later Ground is uploaded to server end.

While any variation occurs for the image content produced, and saves simultaneously uploaded videos, every 500ms (target Time interval) picture is saved, picture is sent to local server in real time, is used for image recognition.

Server immediately completes the image recognition to picture after receiving picture, using AI technology, judge be in image It is no to have object noxious livings, such as mouse, cockroach etc., or only foreign matter such as flies at the non-insect pest invasion scene, that is, enter data Analytical procedure.

(2) data analyze (local server, image recognition) step:

To the image application image recognizer that video monitoring equipment is returned, the knowledge of the worms plagues of rats such as mouse, cockroach is carried out Not.It is true when being identified as, then it is assumed that the moment has found the plague of rats, insect pest, transfers and downloads the period, the worm plague of rats of the camera Video haunt for further analyzing and (receiving when server receives continuous picture collection, and be judged as that damaged by vermin is invaded, in fact When request the video of entire period);When being identified as vacation, then it is assumed that the Dynamic Recognition at the moment is unrelated with the worm plague of rats, do not make into The processing of one step.

Optionally, in order to improve differentiation accuracy rate, manual review can be introduced, is all strictly with what confirmation was detected every time There are mouse, cockroach etc. to haunt, to improve the accuracy rate differentiated to the worm plague of rats.

(3) prompt alarm step (can be used for promptly killing mouse):

When by the identification to pictures, when detecting that mouse haunts, server sends warning message, instruction to output end Dining room operation personnel, pest control personnel take measures, and provide image and video playback, indicate the quilts such as mouse, cockroach The harmful organism identified tentatively judges position and harm that it occurs convenient for operator, and takes timely control measure.

The scene that promptly kills mouse is suitble to computer room, hospital etc. to be impermissible for the monitoring that mouse suffers from the place occurred, Attended mode.It is sending out Indicate that related personnel takes measures immediately after existing mouse feelings, system is responsible for providing picture and video playback in time, for the reference that kills mouse.

Optionally, warning message can also be sent by modes such as short message, pushed informations (in addition to being linked into local net network Except, local server can also be linked into internet).

(4) path analysis step:

By the further analysis to video data, the movement routine of the harmful organisms such as mouse, cockroach is extracted, is marked Invasion point when mouse haunts conceals the information such as point, travelling route, Active duration, skin color, for formulating control mouse, control worm Further embodiment is shown in user terminal.

Mouse path, which is shown, can be used punctuate expression, with a string of digital representations from small to large on line segment, to indicate old The carry out direction of mouse or cockroach.

(5) it shows output and kills mouse, deinsectization suggestion (for conventional worm rat plague control):

Local server summarizes the worm plague of rats information that each monitoring device is collected, according to the historical path that the worm plague of rats haunts, It is suitble to dispose the position of the instruments such as mouse sticking plate, cockroach-killing room in conjunction with place place, the position for providing placement is suggested.

By outlet terminal, dining room operation personnel and pest control personnel are presented to, daily provide report automatically, it is optional Ground, by the optional mode such as public platform, short message, is pushed to dining room operation or related personnel in the environment of having network condition.

Data dimension to show further includes that the worm plague of rats in evening on the previous day/same day enlivens duration, insect pest type, capture Quantity etc..

The above-mentioned technical proposal of alternative embodiment through the invention passes through the localization portion for not depending on broadband network of offer Framework is affixed one's name to, monitoring and the remedial proposal of protection against rodents insect prevention work is provided for indoor spaces such as dining rooms, protection against rodents insect prevention is instructed to work Effectively carry out.

The embodiments of the present invention also provide a kind of storage medium, computer program is stored in the storage medium, wherein The computer program is arranged to execute the step in any of the above-described embodiment of the method when operation.

Optionally, in the present embodiment, above-mentioned storage medium can be set to store by executing based on following steps Calculation machine program:

S1 receives the target image that target monitoring equipment is sent by local net network, wherein target image is target The image that monitoring device is shot in the case where the monitored picture for detecting target monitoring equipment changes;

S2 carries out recongnition of objects to target image;

In the case that S3 in identifying target image includes target object, target identification result is determined, wherein Target identification result is for indicating target object occur in the monitoring area of target monitoring equipment.

Optionally, in the present embodiment, above-mentioned storage medium can be set to store by executing based on following steps Calculation machine program:

S1, monitoring device carry out mobile detection to the monitoring area of monitoring device, wherein mobile detection is for detecting monitoring Whether the monitored picture of equipment changes;

S2, in the case where detecting that monitored picture changes, monitoring device acquisition shoot to monitoring area The target image arrived;

Target image is uploaded to local server by local net network by S3, monitoring device, wherein target image is used for Indicate that local server carries out recongnition of objects to target image.

Optionally, in the present embodiment, above-mentioned storage medium can include but is not limited to: USB flash disk, read-only memory (Read- Only Memory, referred to as ROM), it is random access memory (Random Access Memory, referred to as RAM), mobile hard The various media that can store computer program such as disk, magnetic or disk.

The embodiments of the present invention also provide a kind of electronic device, including memory and processor, stored in the memory There is computer program, which is arranged to run computer program to execute the step in any of the above-described embodiment of the method Suddenly.

Optionally, above-mentioned electronic device can also include transmission device and input-output equipment, wherein the transmission device It is connected with above-mentioned processor, which connects with above-mentioned processor.

Optionally, in the present embodiment, above-mentioned processor can be set to execute following steps by computer program:

S1 receives the target image that target monitoring equipment is sent by local net network, wherein target image is target The image that monitoring device is shot in the case where the monitored picture for detecting target monitoring equipment changes;

S2 carries out recongnition of objects to target image;

In the case that S3 in identifying target image includes target object, target identification result is determined, wherein Target identification result is for indicating target object occur in the monitoring area of target monitoring equipment.

Optionally, in the present embodiment, above-mentioned processor can be set to execute following steps by computer program:

S1, monitoring device carry out mobile detection to the monitoring area of monitoring device, wherein mobile detection is for detecting monitoring Whether the monitored picture of equipment changes;

S2, in the case where detecting that monitored picture changes, monitoring device acquisition shoot to monitoring area The target image arrived;

Target image is uploaded to local server by local net network by S3, monitoring device, wherein target image is used for Indicate that local server carries out recongnition of objects to target image.

Optionally, the specific example in the present embodiment can be with reference to described in above-described embodiment and optional embodiment Example, details are not described herein for the present embodiment.

Obviously, those skilled in the art should be understood that each module of the above invention or each step can be with general Computing device realize that they can be concentrated on a single computing device, or be distributed in multiple computing devices and formed Network on, optionally, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored It is performed by computing device in the storage device, and in some cases, it can be to be different from shown in sequence execution herein Out or description the step of, perhaps they are fabricated to each integrated circuit modules or by them multiple modules or Step is fabricated to single integrated circuit module to realize.In this way, the present invention is not limited to any specific hardware and softwares to combine.

The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field For art personnel, the invention may be variously modified and varied.It is all within principle of the invention, it is made it is any modification, etc. With replacement, improvement etc., should all be included in the protection scope of the present invention.

Claims (31)

1. a kind of monitoring method of target object characterized by comprising
Local server receives the target image that target monitoring equipment is sent by local net network, wherein the target figure As being what the target monitoring equipment was shot in the case where the monitored picture for detecting the target monitoring equipment changes Image;
The local server carries out recongnition of objects to the target image;
In the case where including the target object in identifying the target image, the local server determines target Recognition result, wherein the target identification result is for indicating in the monitoring area of the target monitoring equipment described in appearance Target object.
2. the method according to claim 1, wherein determining the target identification knot in the local server After fruit, the method also includes:
The local server obtains the target video comprising the target image, wherein the target video includes the mesh The video data that shoots in the target time period of mark monitoring device, the target time section be the first moment to the second moment when Between section, first moment be the target monitoring equipment detect the monitored picture start variation at the time of, described second At the time of moment is that the target monitoring equipment detects that the monitored picture terminates variation.
3. according to the method described in claim 2, it is characterized in that,
The local server receives the target image that the target monitoring equipment is sent by the local net network It include: that the local server receives the target monitoring equipment by the local net network in the target time section Every the target image that target time interval is sent;
It includes: more than between the object time that the local server, which obtains the target video comprising the target image, Every time do not receive the target image that the target monitoring equipment is sent in the case where, the local server obtains The target video comprising the target image.
4. according to the method described in claim 2, it is characterized in that, the local server is obtained comprising the target image The target video includes:
The local server sends request message to the target monitoring equipment, wherein the request message is used for request packet Video containing the target image;It receives the target monitoring equipment and responds the target video that the request message is sent; Alternatively,
The video data concentration that the local server is saved from the local server transfers the target video, wherein institute State the video that sets of video data includes the target monitoring equipment shooting that the target monitoring equipment uploads.
5. according to the method described in claim 2, it is characterized in that, obtaining in the local server includes the target image The target video after, the method also includes:
The local server sends a warning message to output end, wherein the warning information is for setting the target monitoring Occur the target object in standby monitoring area to be alerted, the warning information includes at least one of: the target Image, the target video.
6. according to the method described in claim 5, it is characterized in that, in the local server to described in output end transmission Before warning information, the method also includes:
The local server determines target information corresponding with the target video according to the target video, wherein institute Stating target information includes at least one of: travel path of the target object in the monitoring area, the target pair Invasion point as entering the monitoring area, the target object concealing a little in the monitoring area, the target object Activity time in the monitoring area, the characteristics of objects of the target object, the warning information further include: the target Information.
7. according to the method described in claim 6, it is characterized in that, being determined in the local server according to the target video Out after the target information, the method also includes:
The local server marks the target object and corresponding with the target video in the target video The target information, wherein include travel path of the target object in the monitoring area in the target information In the case of, the travel path is by having number at the target object a string on the motion track in the monitoring area The punctuate of word indicates, and the number on the punctuate is sequentially increased along the moving direction of the target object.
8. according to the method described in claim 6, it is characterized in that, being determined in the local server according to the target video Out after target information corresponding with the target video, the method also includes:
In the case where the target information includes travel path of the target object in the monitoring area, the local Server generates prompt information according to the travel path, wherein the prompt information eliminates the target object for prompting Mode, the warning information further includes the prompt information.
9. according to the method described in claim 6, it is characterized in that, in the local server to described in output end transmission Before warning information, the method also includes:
The local server obtains the objective result of multiple monitor videos, wherein the multiple monitor video is target area The monitor video that interior multiple monitoring devices upload, the multiple monitoring device includes the target monitoring equipment, the target It as a result is the target information corresponding with the multiple monitor video;
The local server determines one or more positions in the target area according to the objective result of acquisition Information, wherein one or more of location informations include at least one of: the target object enters the target area Invasion point, the target object concealing a little in the target area, the target object in the target area go out Occurrence number is more than targeted number threshold value and/or time of occurrence is more than the location point of object time threshold value.
10. the method according to claim 1, wherein the local server carries out mesh to the target image Marking Object identifying includes:
In the case where the target image is multiple target video frame images, mesh is carried out to each target video frame image Object detection is marked, the characteristics of image of each target video frame image is obtained, wherein each target video frame image is used There is the object of movement in the monitoring area in instruction, described image feature is used in the object that there is movement, Similarity between the target object is greater than the object region where the object of first threshold;
Motion feature is determined according to the characteristics of image of each target video frame image, wherein
The motion feature is used to indicate the movement velocity that there is the object of movement described in the multiple target video frame image And the direction of motion;
According to the characteristics of image of the motion feature and each target video frame image, the multiple target video frame is determined Whether the target object is had in image.
11. according to the method described in claim 10, it is characterized in that, carrying out target to each target video frame image Object detection, before obtaining the described image feature of each target video frame image, the method also includes:
Obtain the video file that the target monitoring equipment shoots the monitoring area;
Pumping frame sampling is carried out to the video file, obtains one group of video frame images;
It is determined in one group of video frame images according to the pixel value of the pixel in one group of video frame images multiple The target video frame image.
12. according to the method for claim 11, which is characterized in that according to the figure of each target video frame image As feature determines that the motion feature includes:
Target vector corresponding with object region represented by the characteristics of image of each target video frame image is obtained, Obtain multiple target vectors, wherein each target vector is for indicating in a corresponding target video frame image Movement velocity and the direction of motion of the object that there is movement when by the object region;By the multiple target Vector forms first object vector according to time sequencing of each target video frame image in the video file, In, the motion feature includes the first object vector;Alternatively,
Obtain two-dimentional light stream corresponding with object region represented by the characteristics of image of each target video frame image Figure obtains multiple two-dimentional light stream figures, wherein each two-dimentional light stream figure includes a corresponding target video frame image Described in there is movement velocity and the direction of motion of the object of movement when by the object region;By the multiple two It ties up light stream figure and forms three-dimensional second target according to time sequencing of each target video frame image in the video file Vector, wherein the motion feature includes three-dimensional second object vector.
13. according to the method described in claim 10, it is characterized in that, according to the motion feature and each target video The described image feature of frame image determines that whether having the target object in the multiple target video frame image includes:
The characteristics of image of the motion feature and each target video frame image is input to preparatory trained nerve net In network model, Object identifying result is obtained, wherein the Object identifying result is for indicating the multiple target video frame image In whether have the target object.
14. according to the method for claim 13, which is characterized in that by the motion feature and each target video frame The characteristics of image of image is input in preparatory trained neural network model, and obtaining Object identifying result includes:
Each described image feature is passed through into the neural net layer structure including convolutional layer, regularization layer and activation primitive layer, is obtained To multiple first eigenvectors;The multiple first eigenvector is merged with the motion feature, obtains second feature Vector;The second feature vector is input to full articulamentum to classify, obtains the first classification results, wherein the nerve Network model includes the neural net layer structure and the full articulamentum, and the Object identifying result includes first classification As a result, first classification results are for indicating whether have the target object in the multiple target video frame image; Alternatively,
Each described image feature is passed through into the first nerves network layer knot including convolutional layer, regularization layer and activation primitive layer Structure obtains multiple first eigenvectors;By the motion feature pass through including convolutional layer, regularization layer, activation primitive layer the Two neural net layer structures, obtain second feature vector;By the multiple first eigenvector and the second feature vector into Row fusion, obtains third feature vector;The third feature vector is input to full articulamentum to classify, obtains the second classification As a result, wherein the neural network model include the first nerves network layer structure, the nervus opticus network layer structure and The full articulamentum, the Object identifying result include second classification results, and second classification results are for indicating institute It states in multiple target video frame images and whether has the target object.
15. according to claim 1 to method described in any one of 14, which is characterized in that in the local server to described After target image carries out recongnition of objects, the method also includes:
In the case where not including the target object in identifying the target image, the local server deletes the mesh Mark monitoring device upload target video, wherein the target video include the target monitoring equipment in the target time period The video data of shooting, the target time section are the period at the first moment to the second moment, and first moment is described At the time of target monitoring equipment detects that the monitored picture starts variation, second moment is target monitoring equipment inspection At the time of measuring the monitored picture terminates variation.
16. a kind of monitoring method of target object characterized by comprising
Monitoring device carries out mobile detection to the monitoring area of the monitoring device, wherein the mobile detection is for detecting Whether the monitored picture for stating monitoring device changes;
In the case where detecting that the monitored picture changes, the monitoring area is clapped in the monitoring device acquisition The target image taken the photograph;
The target image is uploaded to local server by local net network by the monitoring device, wherein the target figure Recongnition of objects is carried out to the target image as being used to indicate the local server.
17. according to the method for claim 16, which is characterized in that the monitoring device passes through the target image described Local net network is uploaded to the local server
The target that the monitoring device takes the monitoring device every target time interval in the target time period Image is uploaded to the local server, wherein the target time section is the period at the first moment to the second moment, described At the time of first moment was that the monitoring device detects that the monitored picture starts variation, second moment is the monitoring At the time of equipment detects that the monitored picture terminates variation.
18. according to the method for claim 16, which is characterized in that the target image is passed through institute in the monitoring device Local net network is stated to be uploaded to after the local server, the method also includes:
In the case where detecting that the monitored picture terminates variation, target video is uploaded to the local by the monitoring device Server, wherein the target video includes the video data that the monitoring device is shot in the target time period, the target Period is the period at the first moment to the second moment, and first moment is that the monitoring device detects the monitoring picture At the time of face starts variation, at the time of second moment is that the monitoring device detects that the monitored picture terminates variation.
19. according to the method for claim 18, which is characterized in that be uploaded to target video in the monitoring device described Before local server, the method also includes:
The monitoring device determines first moment and second moment;
The monitoring device reads out the target video from the memory of the monitoring device, wherein the target video is The video data at third moment to the 4th moment of monitoring device shooting, the third moment be first moment it Before, at the time of be separated by first time period with first moment, the 4th moment be second moment after, with described the At the time of two moment were separated by the second time interval.
20. according to the method for claim 18, which is characterized in that the target video is uploaded to institute by the monitoring device Before stating local server, the method also includes:
In the case where detecting that the monitored picture terminates variation, the monitoring device is read from the memory of the monitoring device Take out the target video;
The target video is written in the storage unit of the monitoring device by the monitoring device.
21. according to the method for claim 18, which is characterized in that the target video is uploaded to institute by the monitoring device Stating local server includes:
During the target video is uploaded to the local server, the monitoring device detects the local area network Network is in abnormality;
The monitoring device saves the upload progress of the target video;
In the case where detecting that the local net network is in normal condition, the monitoring device is according to the upload of preservation Progress continues to upload the target video to the local server.
22. 6 to 21 described in any item methods according to claim 1, which is characterized in that the monitoring device sets the monitoring The standby monitoring area carries out mobile detection
The monitoring device moves the monitoring area of the monitoring device by the camera of the monitoring device Detection.
23. a kind of monitoring system of target object characterized by comprising target monitoring equipment and local server, the mesh Mark monitoring device is connected with the local server by local net network, wherein
The target monitoring equipment carries out mobile detection for the monitoring area to the target monitoring equipment, wherein the shifting Dynamic detection is for detecting whether the monitored picture of the target monitoring equipment changes;Detecting the monitored picture generation In the case where variation, the target image shot to the monitoring area is obtained;The target image is passed through described Local net network is uploaded to the local server;
The local server, the target sent for receiving the target monitoring equipment by the local net network Image;Recongnition of objects is carried out to the target image;It include the target object in identifying the target image In the case where, determine target identification result, wherein the target identification result is for indicating occur in the monitoring area The target object.
24. system according to claim 23, which is characterized in that the target monitoring equipment is also used to:
In the target time period, the target monitoring equipment is uploaded every the target image that target time interval takes To the local server, wherein the target time section is the period at the first moment to the second moment, first moment At the time of detecting that the monitored picture starts variation for the monitoring device, second moment is monitoring device detection At the time of terminating variation to the monitored picture.
25. system according to claim 23, which is characterized in that the target monitoring equipment is also used to:
In the case where detecting that the monitored picture terminates variation, target video is uploaded to the local server, wherein The target video includes the video data that the monitoring device is shot in the target time period, and the target time section is first The period at moment to the second moment, first moment are that the target monitoring equipment detects that the monitored picture starts to become At the time of change, at the time of second moment is that the target monitoring equipment detects that the monitored picture terminates variation.
26. system according to claim 25, which is characterized in that the system also includes output end, the output end with The local server is connected, wherein
The local server is also used to receive the target video that the target monitoring equipment uploads;According to the target Video determines target information corresponding with the target video, wherein the target information includes at least one of: described Travel path of the target object in the monitoring area, the invasion point into the monitoring area of the target object, institute State target object concealing a little in the monitoring area, activity time of the target object in the monitoring area, institute State the characteristics of objects of target object;It include travel path of the target object in the monitoring area in the target information In the case where, prompt information is generated according to the travel path, wherein the prompt information eliminates the target pair for prompting The mode of elephant;It sends a warning message to the output end, wherein the warning information is used for the prison to the target monitoring equipment There is the target object in control region to be alerted, the warning information includes at least one of: the target image, institute State target video, the target information, the prompt information;
The output end, for receiving the warning information;At least one of is shown on the screen of the output end: described Target image, the target video, the target information, the prompt information.
27. the system according to any one of claim 23 to 26, which is characterized in that the local server is also used to:
In the case where not including the target object in identifying the target image, deletes the target monitoring equipment and upload Target video, wherein the target video includes the video data that shoots in the target time period of the target monitoring equipment, The target time section is the period at the first moment to the second moment, and first moment is target monitoring equipment detection At the time of starting variation to the monitored picture, second moment is that the target monitoring equipment detects the monitored picture At the time of terminating variation.
28. a kind of monitoring device of target object is applied to local server characterized by comprising
Receiving module, the target image sent for receiving target monitoring equipment by local net network, wherein the target Image is that the target monitoring equipment is shot in the case where the monitored picture for detecting the target monitoring equipment changes Image;
Identification module, for carrying out recongnition of objects to the target image;
Determining module determines target in the case where for including the target object in identifying the target image Recognition result, wherein the target identification result is for indicating in the monitoring area of the target monitoring equipment described in appearance Target object.
29. a kind of monitoring device of target object is applied to monitoring device characterized by comprising
Detection module carries out mobile detection for the monitoring area to the monitoring device, wherein the mobile detection is for examining Whether the monitored picture for surveying the monitoring device changes;
Module is obtained, for obtaining and carrying out to the monitoring area in the case where detecting that the monitored picture changes Shoot obtained target image;
Uploading module, for the target image to be uploaded to local server by local net network, wherein the target figure Recongnition of objects is carried out to the target image as being used to indicate the local server.
30. a kind of storage medium, which is characterized in that be stored with computer program in the storage medium, wherein the computer Program is arranged to execute method described in any one of described claim 1 to 22 when operation.
31. a kind of electronic device, including memory and processor, which is characterized in that be stored with computer journey in the memory Sequence, the processor are arranged to run the computer program to execute described in any one of described claim 1 to 22 Method.
CN201910069792.0A 2019-01-24 2019-01-24 Target object monitoring method, device and system CN109922310B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910069792.0A CN109922310B (en) 2019-01-24 2019-01-24 Target object monitoring method, device and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910069792.0A CN109922310B (en) 2019-01-24 2019-01-24 Target object monitoring method, device and system

Publications (2)

Publication Number Publication Date
CN109922310A true CN109922310A (en) 2019-06-21
CN109922310B CN109922310B (en) 2020-11-17

Family

ID=66960716

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910069792.0A CN109922310B (en) 2019-01-24 2019-01-24 Target object monitoring method, device and system

Country Status (1)

Country Link
CN (1) CN109922310B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110677588A (en) * 2019-10-14 2020-01-10 浙江大华技术股份有限公司 Picture acquisition method and device, storage medium and electronic device
CN110996067A (en) * 2019-12-19 2020-04-10 哈尔滨融智爱科智能科技有限公司 Personnel safety real-time intelligent video monitoring system under high-risk operation environment based on deep learning

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101271527A (en) * 2008-02-25 2008-09-24 北京理工大学 Exception action detecting method based on athletic ground partial statistics characteristic analysis
CN101739550A (en) * 2009-02-11 2010-06-16 北京智安邦科技有限公司 Method and system for detecting moving objects
CN102148959A (en) * 2010-02-09 2011-08-10 北京中星微电子有限公司 Video monitoring system and method for detecting moving target of image thereof
CN102202216A (en) * 2011-05-13 2011-09-28 南京信息工程大学 Mouse and insect monitoring device and method for performing image tracking and identification by adopting same
CN102819764A (en) * 2012-07-18 2012-12-12 郑州金惠计算机系统工程有限公司 Method for counting pedestrian flow from multiple views under complex scene of traffic junction
CN104933416A (en) * 2015-06-26 2015-09-23 复旦大学 Micro expression sequence feature extracting method based on optical flow field
CN105279898A (en) * 2015-10-28 2016-01-27 小米科技有限责任公司 Alarm method and device
WO2018106608A1 (en) * 2016-12-05 2018-06-14 Ring Inc. Passing vehicle filters for audio/video recording and communication devices
CN108259830A (en) * 2018-01-25 2018-07-06 深圳冠思大数据服务有限公司 Mouse based on Cloud Server suffers from intelligent monitor system and method

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101271527A (en) * 2008-02-25 2008-09-24 北京理工大学 Exception action detecting method based on athletic ground partial statistics characteristic analysis
CN101739550A (en) * 2009-02-11 2010-06-16 北京智安邦科技有限公司 Method and system for detecting moving objects
CN102148959A (en) * 2010-02-09 2011-08-10 北京中星微电子有限公司 Video monitoring system and method for detecting moving target of image thereof
CN102202216A (en) * 2011-05-13 2011-09-28 南京信息工程大学 Mouse and insect monitoring device and method for performing image tracking and identification by adopting same
CN102819764A (en) * 2012-07-18 2012-12-12 郑州金惠计算机系统工程有限公司 Method for counting pedestrian flow from multiple views under complex scene of traffic junction
CN104933416A (en) * 2015-06-26 2015-09-23 复旦大学 Micro expression sequence feature extracting method based on optical flow field
CN105279898A (en) * 2015-10-28 2016-01-27 小米科技有限责任公司 Alarm method and device
WO2018106608A1 (en) * 2016-12-05 2018-06-14 Ring Inc. Passing vehicle filters for audio/video recording and communication devices
CN108259830A (en) * 2018-01-25 2018-07-06 深圳冠思大数据服务有限公司 Mouse based on Cloud Server suffers from intelligent monitor system and method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
张作楠: "机器人视觉伺服跟踪系统的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110677588A (en) * 2019-10-14 2020-01-10 浙江大华技术股份有限公司 Picture acquisition method and device, storage medium and electronic device
CN110996067A (en) * 2019-12-19 2020-04-10 哈尔滨融智爱科智能科技有限公司 Personnel safety real-time intelligent video monitoring system under high-risk operation environment based on deep learning

Also Published As

Publication number Publication date
CN109922310B (en) 2020-11-17

Similar Documents

Publication Publication Date Title
US9830511B2 (en) Method of searching data to identify images of an object captured by a camera system
CN105447459B (en) A kind of unmanned plane detects target and tracking automatically
CA2973866C (en) Video identification and analytical recognition system
CN104919794B (en) For extracting the method and system of metadata from master-slave mode camera tracking system
US9471832B2 (en) Human activity determination from video
Qureshi et al. Machine vision for counting fruit on mango tree canopies
Beery et al. Recognition in terra incognita
AU2012340862B2 (en) Geographic map based control
AU2012355879B2 (en) Cloud-based video surveillance management system
EP2219379B1 (en) Social network construction based on data association
Fleck et al. Smart camera based monitoring system and its application to assisted living
US7760908B2 (en) Event packaged video sequence
Haering et al. The evolution of video surveillance: an overview
KR100905504B1 (en) Video tripwire
US8620028B2 (en) Behavioral recognition system
US10152858B2 (en) Systems, apparatuses and methods for triggering actions based on data capture and characterization
CN106203458B (en) Crowd video analysis method and system
US7840515B2 (en) System architecture and process for automating intelligent surveillance center operations
EP3745713A1 (en) Cloud server-based rodent outbreak smart monitoring system and method
Boissard et al. A cognitive vision approach to early pest detection in greenhouse crops
CN105378752B (en) The on-line study system for detecting and counting for people
US8438175B2 (en) Systems, methods and articles for video analysis reporting
CA2541437C (en) System and method for searching for changes in surveillance video
KR20180135898A (en) Systems and methods for training object classifiers by machine learning
US6900729B2 (en) Thermal signature intensity alarmer

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant