CN114049681A - Monitoring method, identification method, related device and system - Google Patents
Monitoring method, identification method, related device and system Download PDFInfo
- Publication number
- CN114049681A CN114049681A CN202111170226.2A CN202111170226A CN114049681A CN 114049681 A CN114049681 A CN 114049681A CN 202111170226 A CN202111170226 A CN 202111170226A CN 114049681 A CN114049681 A CN 114049681A
- Authority
- CN
- China
- Prior art keywords
- target
- identification
- target person
- image sequence
- monitoring
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The embodiment of the application discloses a monitoring method, an identification method, a related device and a system, wherein the monitoring method comprises the following steps: carrying out first identification on a monitoring video to obtain a target video frame sequence; determining a target region humanoid image sequence according to the target video frame sequence; uploading the target area human-shaped image sequence to a recognition device, wherein the target area human-shaped image sequence is used for determining whether the target person is included in the target area human-shaped image sequence or not through second recognition by the recognition device, and the recognition accuracy of the second recognition is greater than that of the first recognition. By adopting the technical scheme provided by the embodiment of the application, the human-shaped image sequence of the target area obtained after the first identification is uploaded to the identification equipment, and the human-shaped image sequence is much smaller than the surveillance video, so that the data volume transmitted to the identification equipment is reduced, the identification burden of the identification equipment is lightened, and the identification of the identification equipment on the target person is accelerated.
Description
Technical Field
The present application relates to the field of monitoring, and in particular, to a monitoring method, an identification method, a related device and a related system.
Background
Gait is a way of walking of people, can be used for identification, and has the characteristic of remote identification relative to biological characteristics such as fingerprints, human faces or irises which can be identified only by contact or close distance, so the gait identification is an important research direction in the monitoring field.
When gait recognition is carried out, the monitoring device collects videos and then sends the videos to recognition equipment, such as a cloud server, and the cloud server carries out gait recognition on the received videos. Because the gait is a dynamic process and the algorithm complexity of gait recognition is high, the gait recognition needs to be realized by using a cloud server with higher performance.
Because the amount of video data transmitted to the cloud server by the monitoring device is large, the transmission process takes long time, and the time spent by the cloud server in identifying the received video is also long, the gait identification of the video by the cloud server has large time delay, so that the monitoring method provided by the prior art is not timely enough for timely discovering time-delay-sensitive monitoring scenes such as criminal suspects and the like according to the monitoring video.
Disclosure of Invention
The application provides a monitoring method, an identification method, a related device and a related system, which can reduce data transmission quantity and accelerate identification of a target person.
In a first aspect, an embodiment of the present application provides a monitoring method, where the method includes:
performing first identification screening processing on a monitored video, and selecting video frames including human shapes in the monitored video to obtain a first video frame sequence;
extracting a human-shaped contour image from each video frame in the first video frame sequence to obtain a gait energy map;
performing first identification on the gait energy map to obtain the similarity between the gait energy map and a preset gait energy map of the target person;
when the similarity is greater than a second threshold, determining the first sequence of video frames as the sequence of target video frames.
The sequence of target video frames comprising a plurality of target video frames;
determining a target region humanoid image sequence according to the target video frame sequence, wherein the target region humanoid image sequence comprises a region which is possibly provided with a target person in at least one target video frame in the plurality of target video frames;
uploading the target area human-shaped image sequence to a recognition device, wherein the target area human-shaped image sequence is used for determining whether the target person is included in the target area human-shaped image sequence or not through second recognition by the recognition device, and the recognition accuracy of the second recognition is greater than that of the first recognition.
According to the technical scheme, the target region human-shaped image sequence obtained after the first recognition is uploaded to the recognition device and is much smaller than the surveillance video, so that the data volume transmitted to the recognition device is reduced, the recognition burden of the recognition device is reduced, and the recognition of the recognition device on the target person is accelerated.
Based on the first aspect, in some possible embodiments of the present application, the first identifying a surveillance video to obtain a target video frame sequence includes:
performing the first identification on each video frame in the monitored video to obtain the probability that each video frame in the monitored video comprises the target person;
and composing the target video frame sequence by the video frames with the corresponding probability larger than a first preset threshold value.
According to the technical scheme provided by the embodiment of the application, the target video frame sequence is limited, in the embodiment, each video frame in the monitoring video is identified, the probability that each video frame comprises a target person is obtained, and the target video frame sequence is formed by the video frames with the corresponding probability being larger than a first preset threshold value.
In the embodiment, the target video frame sequence is determined by using the similarity between a gait energy map obtained by a monitoring video and a preset gait energy map of a target person.
In some possible embodiments of the present application based on the first aspect, the method further includes:
acquiring the appearance characteristics of the target person identified by the identification equipment;
the first identification of the monitoring video to obtain a target video frame sequence comprises the following steps:
and performing the first identification on the monitoring video according to the appearance characteristics to obtain the target video frame sequence.
According to the technical scheme provided by the embodiment of the application, the appearance characteristics of the target person obtained by identification equipment are utilized to identify the monitored video, and the appearance characteristics can be as follows: clothing characteristics, accessory characteristics, age information, gender information, or hair style characteristics, etc. The appearance characteristics of the target person obtained by the identification equipment are used for first identification, so that the identification accuracy can be improved, the data volume uploaded to the identification equipment is reduced, the identification burden of the identification equipment is reduced, and the identification efficiency of the identification equipment is improved.
In some possible embodiments of the present application based on the first aspect, the method further includes:
acquiring a tracking instruction sent by the identification equipment;
and adjusting the monitoring angle according to the tracking instruction.
The technical scheme provided by the embodiment of the application is beneficial to tracking and monitoring the target object.
In a second aspect, an embodiment of the present application provides an identification method, where the method includes:
acquiring a target area human-shaped image sequence uploaded by a monitoring device; the target region humanoid image sequence is determined by a target video frame sequence, the target video frame sequence comprises a plurality of target video frames, the target video frame sequence is obtained by performing first identification on a monitoring video through the monitoring device, and the target region humanoid image sequence comprises a region which possibly comprises a target person in at least one target video frame in the plurality of target video frames;
and performing second recognition on the target region human-shaped image sequence based on the pre-stored characteristics of the target person, and determining whether the target region human-shaped image sequence comprises the target person, wherein the recognition accuracy of the second recognition is greater than that of the first recognition.
According to the technical scheme, the identification object of the identification equipment is the target area human-shaped image sequence uploaded by the monitoring device, and is much smaller than the monitoring video, so that the transmitted data volume is reduced, the identification burden of the identification equipment is reduced, and the identification of the identification equipment on a target person is accelerated.
In some possible embodiments of the present application based on the second aspect, the method further includes:
and sending a tracking instruction to the monitoring device, wherein the tracking instruction is used for instructing the monitoring device to adjust a monitoring angle.
The technical scheme provided by the embodiment of the application is beneficial to tracking and monitoring the target object.
In some possible embodiments of the present application based on the second aspect, the pre-stored characteristics of the target person include one or more of the following characteristics: a gait energy map of the target person, facial features of the target person, anthropomorphic features of the target person, skeletal features of the target person, and optical flow features of the target person.
Based on the second aspect, in some possible embodiments of the application, when it is determined that the target person is included in the target area human-shaped image sequence, the method further includes:
extracting motion characteristics of the target person from the target area humanoid image sequence, wherein the motion characteristics comprise a moving direction;
predicting the activity area of the target person according to the motion characteristics;
sending a monitoring instruction including characteristics of the target person to a monitoring device in the activity area.
The technical scheme provided by the embodiment of the application is beneficial to linkage monitoring of the target object.
Based on the second aspect, in some possible embodiments of the present application, when it is determined that the target person is included in the target region human-shaped image sequence, the method further includes:
extracting appearance characteristics of the target person from the target area humanoid image sequence;
and sending the appearance characteristics of the target person to the monitoring device, wherein the appearance characteristics of the target person are used for the monitoring device to perform the first identification on the monitored video.
According to the technical scheme provided by the embodiment of the application, when the target character is determined to be included in the target region humanoid image sequence, the appearance characteristic of the target character is extracted from the target region humanoid image sequence, and the appearance characteristic of the target character is sent to the monitoring equipment for first recognition, wherein the appearance characteristic can be that: clothing characteristics, accessory characteristics, age information, gender information, or hair style characteristics, etc. The accuracy of identification can be improved by utilizing the appearance characteristics of the target person to perform first identification, the data volume uploaded to the identification equipment is reduced, the identification burden of the identification equipment is reduced, and the identification efficiency of the identification equipment is improved.
Based on the second aspect, in some possible embodiments of the present application, when the pre-stored features of the target person include two or more features, the second identifying the target area human-shaped image sequence based on the pre-stored features of the target person includes:
and extracting two or more pre-stored characteristics of the target figure by utilizing the target area figure image sequence, and performing the second identification on the target area figure image sequence based on the extracted two or more pre-stored characteristics of the target figure.
The technical scheme provided by the embodiment of the application limits the implementation of the second identification.
In a third aspect, an embodiment of the present application provides a monitoring apparatus, where the monitoring apparatus includes:
the first identification unit is used for carrying out first identification processing on a monitored video, screening out video frames including human shapes in the monitored video and obtaining a first video frame sequence; extracting a human-shaped contour image from each video frame in the first video frame sequence to obtain a gait energy map; performing first identification on the gait energy map to obtain the similarity between the gait energy map and a preset gait energy map of the target person; when the similarity is greater than a second threshold, determining the first sequence of video frames as the sequence of target video frames.
The sequence of target video frames comprising a plurality of target video frames;
a first processing unit, configured to determine a target region humanoid image sequence according to the target video frame sequence, where the target region humanoid image sequence includes a region that may include a target person in at least one target video frame of the multiple target video frames;
the first uploading unit is used for uploading the target area human-shaped image sequence to the recognition equipment, the target area human-shaped image sequence is used for determining whether the target area human-shaped image sequence comprises the target person or not through second recognition by the recognition equipment, and the recognition accuracy of the second recognition is greater than that of the first recognition.
According to the technical scheme, the target region human-shaped image sequence obtained after the first recognition is uploaded to the recognition device and is much smaller than the surveillance video, so that the data volume transmitted to the recognition device is reduced, the recognition burden of the recognition device is reduced, and the recognition of the recognition device on the target person is accelerated.
In some possible embodiments of the present application, based on the third aspect,
the first identification unit is specifically configured to perform the first identification on each video frame in the surveillance video to obtain a probability that each video frame in the surveillance video includes the target person; and composing the target video frame sequence by the video frames with the corresponding probability larger than a first preset threshold value.
According to the technical scheme provided by the embodiment of the application, the target video frame sequence is limited, in the embodiment, each video frame in the monitoring video is identified, the probability that each video frame comprises a target person is obtained, and the target video frame sequence is formed by the video frames with the corresponding probability being larger than a first preset threshold value.
Based on the third aspect, in some possible embodiments of the present application, the first identifying unit is specifically configured to screen out video frames including human shapes in the monitored video to obtain a first video frame sequence; extracting a human-shaped contour image from each video frame in the first video frame sequence to obtain a gait energy map; performing first identification on the gait energy map to obtain the similarity between the gait energy map and a preset gait energy map of the target person; when the similarity is greater than a second threshold, determining the first sequence of video frames as the sequence of target video frames.
In the embodiment, the target video frame sequence is determined by using the similarity between a gait energy map obtained by a monitoring video and a preset gait energy map of a target person.
In some possible embodiments of the third aspect, the monitoring apparatus further includes:
a first obtaining unit, configured to obtain appearance characteristics of the target person identified by the identification device;
the first identification unit is used for performing first identification on the monitored video to obtain a target video frame sequence, and is specifically used for performing first identification on the monitored video by using the appearance characteristics of the target person.
According to the technical scheme provided by the embodiment of the application, the appearance characteristics of the target person obtained by identification equipment are utilized to identify the monitored video, and the appearance characteristics can be as follows: clothing characteristics, accessory characteristics, age information, gender information, or hair style characteristics, etc. The appearance characteristics of the target person obtained by the identification equipment are used for first identification, so that the identification accuracy can be improved, the data volume uploaded to the identification equipment is reduced, the identification burden of the identification equipment is reduced, and the identification efficiency of the identification equipment is improved.
Based on the third aspect, in some possible embodiments of the present application, the first obtaining unit is further configured to obtain a tracking instruction sent by the identification device;
the monitoring device further comprises:
and the first adjusting unit is used for adjusting the monitored angle according to the tracking instruction.
The technical scheme provided by the embodiment of the application is beneficial to tracking and monitoring the target object.
The technical scheme provided by the embodiment of the application is beneficial to tracking and monitoring the target object.
In a fourth aspect, an embodiment of the present application provides an identification device, where the identification device includes:
the second acquisition unit is used for acquiring a target area human-shaped image sequence uploaded by the monitoring device; the target region humanoid image sequence is determined by a target video frame sequence, the target video frame sequence comprises a plurality of target video frames, the target video frame sequence is obtained by performing first identification on a monitoring video through the monitoring device, and the target region humanoid image sequence comprises a region which possibly comprises a target person in at least one target video frame in the plurality of target video frames;
and the second identification unit is used for carrying out second identification on the target area human-shaped image sequence based on the pre-stored characteristics of the target person and determining whether the target area human-shaped image sequence comprises the target person, and the identification accuracy of the second identification is greater than that of the first identification.
According to the technical scheme, the identification object of the identification equipment is the target area human-shaped image sequence uploaded by the monitoring device, and is much smaller than the monitoring video, so that the transmitted data volume is reduced, the identification burden of the identification equipment is reduced, and the identification of the identification equipment on a target person is accelerated.
Based on the fourth aspect, in some possible embodiments of the present application, the identification device further includes:
the monitoring device comprises a first sending unit, a second sending unit and a control unit, wherein the first sending unit is used for sending a tracking instruction to the monitoring device, and the tracking instruction is used for indicating the monitoring device to adjust a monitoring angle.
The technical scheme provided by the embodiment of the application is beneficial to tracking and monitoring the target object.
In some possible embodiments of the fourth aspect, the pre-stored characteristics of the target person include one or more of the following characteristics: a gait energy map of the target person, facial features of the target person, anthropomorphic features of the target person, skeletal features of the target person, and optical flow features of the target person.
Based on the fourth aspect, in some possible embodiments of the present application, the identification device may further include:
the second processing unit is used for extracting the motion characteristics of the target person from the target area human-shaped image sequence when the second identification unit determines that the target person is included in the target area human-shaped image sequence, and the motion characteristics comprise the moving direction; predicting the activity area of the target person according to the motion characteristics;
a first sending unit, configured to send a monitoring instruction including characteristics of the target person to a monitoring apparatus in the activity area.
The technical scheme provided by the embodiment of the application is beneficial to linkage monitoring of the target object.
Based on the fourth aspect, in some possible embodiments of the present application, the identification device may further include:
the second processing unit is used for extracting the appearance characteristics of the target person from the target area human-shaped image sequence when the second identification unit determines that the target person is included in the target area human-shaped image sequence;
and the first sending unit is used for sending the appearance characteristics of the target person to the monitoring device, and the appearance characteristics of the target person are used for the monitoring device to perform first identification on the monitored video.
According to the technical scheme provided by the embodiment of the application, when the target character is determined to be included in the target region humanoid image sequence, the appearance characteristic of the target character is extracted from the target region humanoid image sequence, and the appearance characteristic of the target character is sent to the monitoring equipment for first recognition, wherein the appearance characteristic can be that: clothing characteristics, accessory characteristics, age information, gender information, or hair style characteristics, etc. The accuracy of identification can be improved by utilizing the appearance characteristics of the target person to perform first identification, the data volume uploaded to the identification equipment is reduced, the identification burden of the identification equipment is reduced, and the identification efficiency of the identification equipment is improved.
Based on the fourth aspect, in some possible embodiments of the present application, the second identification unit is specifically configured to extract two or more pre-stored features of the target person based on the pre-stored features of the target person when performing the second identification on the target region figure image sequence based on the pre-stored features of the target person, and identify the target region figure image sequence based on the extracted two or more pre-stored features of the target person.
The technical scheme provided by the embodiment of the application limits the implementation method of the second identification.
In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, where the storage medium stores a program, and when the program runs, the monitoring method described in the first aspect or any possible implementation manner of the first aspect of the present application is implemented.
In a sixth aspect, the present application provides a computer-readable storage medium, where the storage medium stores a program, and when the program runs, the identification method described in the second aspect or any possible implementation manner of the second aspect of the present application is implemented.
In a seventh aspect, an embodiment of the present application provides a monitoring apparatus, including: the device comprises a camera, a communication unit, a processor, a memory and a bus; wherein the content of the first and second substances,
the camera is used for acquiring a monitoring video;
the communication unit is used for communicating with the identification equipment;
the camera, the communication unit, the processor and the memory are connected through the bus and complete mutual communication;
the memory stores executable program code;
the processor executes a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to perform the monitoring method according to the first aspect of the present application or any possible implementation manner of the first aspect.
In an eighth aspect, an embodiment of the present application provides an identification device, including: a communication unit, a memory, a processor and a bus; wherein the content of the first and second substances,
the communication unit is used for communicating with the monitoring device;
the memory stores executable program code;
the processor executes a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to execute the identification method of the second aspect of the present application or any possible implementation manner of the second aspect.
In a ninth aspect, an embodiment of the present application provides a monitoring system, which includes a monitoring apparatus and an identification device, where the monitoring apparatus is the monitoring apparatus according to the seventh aspect of the present application, and the identification device is the identification device according to the eighth aspect of the present application.
According to the technical scheme, the target area human-shaped image sequence obtained after the first identification is uploaded to the identification device by the monitoring device and is much smaller than the monitoring video, so that the data volume transmitted to the identification device by the monitoring device is reduced, the identification burden of the identification device is reduced, and the identification of the target person by the identification device is accelerated.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments or the background art of the present application, the drawings required to be used in the embodiments or the background art of the present application will be described below.
Fig. 1 is a schematic view of an application scenario architecture of a monitoring system according to an embodiment of the present application.
Fig. 2A is an interaction flow diagram of a monitoring method according to an embodiment of the present application.
Fig. 2B is a schematic flowchart of obtaining a target video frame sequence according to an embodiment of the present application.
Fig. 2C is a schematic flowchart of obtaining a target video frame sequence according to an embodiment of the present application.
Fig. 3A is a schematic functional structure diagram of a monitoring device according to an embodiment of the present application.
Fig. 3B is a functional structure diagram of a monitoring device according to another embodiment of the present application.
Fig. 3C is a functional structure diagram of a monitoring device according to another embodiment of the present application.
Fig. 4A is a functional structure schematic diagram of an identification device according to an embodiment of the present application.
Fig. 4B is a functional structure diagram of an identification device according to another embodiment of the present application.
Fig. 4C is a functional structure diagram of an identification device according to another embodiment of the present application.
Fig. 5 is a schematic structural diagram of a monitoring device according to an embodiment of the present application.
Fig. 6 is a schematic structural diagram of an identification device according to an embodiment of the present application.
Detailed Description
The terminology used in the description of the embodiments of the present application is for the purpose of describing particular embodiments of the present application only and is not intended to be limiting of the present application.
Fig. 1 is a schematic view of an application scenario architecture of a monitoring system according to an embodiment of the present application, and as shown in fig. 1, the monitoring system 100 includes: a monitoring device 101 and an identification device 103, wherein the monitoring device 101 performs data transmission with the identification device 103 through a network 102. The monitoring device 101 is configured to acquire a monitoring video, and perform first identification on the acquired monitoring video to obtain a target video frame sequence, where the target video frame sequence includes a plurality of target video frames; the monitoring device 101 is further configured to determine a target region humanoid image sequence according to the target video frame sequence, where the target region humanoid image sequence includes a region that may include a target person in at least one target video frame of the plurality of target video frames; the monitoring device 101 uploads the target area human-shaped image sequence to the recognition equipment 103 through the network 102, the recognition equipment 103 performs second recognition on the target area human-shaped image sequence based on the pre-stored characteristics of the target person, and determines whether the target area human-shaped image sequence includes the target person, wherein the recognition accuracy of the second recognition is greater than that of the first recognition.
By adopting the technical scheme provided by the embodiment of the application, the human-shaped image sequence of the target area obtained after the first identification is uploaded to the identification equipment, and the human-shaped image sequence is much smaller than the surveillance video, so that the data volume transmitted to the identification equipment is reduced, the identification burden of the identification equipment is lightened, and the identification of the identification equipment on the target person is accelerated.
Referring to fig. 2A, fig. 2A is an interaction flow diagram of a monitoring method according to an embodiment of the present application, and as shown in fig. 2A, the monitoring method according to an embodiment of the present application may include the following:
201. the monitoring device obtains a monitoring video.
In some possible embodiments of the present application, the monitoring apparatus may obtain the monitoring instruction from the identification device, and obtain the monitoring video according to the monitoring instruction. The monitoring instructions may include characteristics of the target person. The method is used for the monitoring device to identify the acquired monitoring video.
202. The method comprises the steps of carrying out first identification on an obtained monitoring video to obtain a target video frame sequence, wherein the target video frame sequence comprises a plurality of target video frames.
In a possible embodiment of the present application, when the step 202 performs the first identification on the obtained surveillance video to obtain the target video frame sequence, a first identification method for each video frame in the surveillance video may be adopted, as shown in fig. 2B, when the obtained surveillance video is identified, the following steps 2021 to 2022 may be included: wherein the content of the first and second substances,
2021. and performing first identification on each video frame in the monitored video to obtain the probability that each video frame in the monitored video comprises the target person.
2022. And composing the target video frame sequence by the video frames with the corresponding probability larger than a first preset threshold value.
In another possible embodiment of the present application, when the step 202 performs the first recognition on the obtained surveillance video to obtain the target video sequence, the step 202 may perform recognition on the obtained surveillance video in a gait recognition manner, as shown in fig. 2C, when the step 202 performs recognition on the obtained surveillance video, the steps 2023 to 2026 may be included,
2023. and screening out video frames including human shapes in the monitored video to obtain a first video frame sequence.
In some possible embodiments, a human-shape detection model may be used to detect whether each video frame in a surveillance video includes a human shape, and a first video frame is obtained from the video frame including the human shape in the surveillance video, specifically, the human-shape detection model may be flexibly selected according to the hardware configuration of the surveillance device, and in order to ensure the real-time performance of the processing, the human-shape detection model may use a simple image processing method, for example, a deformable component model (DPM), that is, a model obtained by training may be used to perform target matching classification; or a miniaturized or compressed Convolutional Neural Network (CNN) model is adopted for realization, namely, the characteristics in the graph are automatically extracted through a series of convolution and nonlinear activation transformation operations, regression calculation is carried out on the figure position, and end-to-end human shape detection is realized. Considering that the computational resources of the monitoring device are limited, the convolutional neural network can be cut and compressed correspondingly so as to reduce the parameter quantity and the computational complexity of the model.
2024. And extracting a human-shaped contour image from each video frame in the first video frame sequence to obtain a gait energy map.
In some possible embodiments of the present application, a two-value contour map of a human shape may be extracted by using a conventional image processing method or a compressed CNN human shape segmentation model method, and a gait energy map (GEI) may be obtained by adding pixel values corresponding to a plurality of contour maps.
2025. And carrying out first identification on the gait energy diagram to obtain the similarity between the gait energy diagram and a preset gait energy diagram of a target person.
In some possible embodiments of the present application, a CCA (canonical correlation analysis) algorithm may be used to perform projection-based matching on the GEI obtained in step 2024 and a preset GEI feature template of the target person (finding projection deformation of each of the two sets of data, maximizing a similarity value of the two sets of data), calculating a similarity value, and comparing the similarity value with a preset threshold.
The CCA algorithm can be solved by solving the problem of the generalized characteristic value, and is simple and efficient to realize. By setting a threshold, most of the video data not containing the target person can be filtered out on the side of the monitoring device.
2026. When the similarity is greater than a second threshold, determining the first sequence of video frames as the sequence of target video frames.
203. And determining a target region humanoid image sequence according to the target video frame sequence.
Wherein the target region humanoid image sequence comprises a region which is possibly provided with a target person in at least one target video frame in the plurality of target video frames;
in some possible embodiments of the present application, when it is identified in step 202 that the surveillance video does not include the target video frame sequence, the surveillance video is directly discarded and is not processed. When the recognition result in step 202 is that the target video frame sequence is included, the rectangular frame position is obtained according to a human shape detection algorithm, and the original video frame is cut to obtain a rectangular region including a human shape, that is, a human shape image of a target Region (ROI), so that a human shape image sequence of the target region can be obtained.
204. And uploading the target area human-shaped image sequence to a recognition device.
In some possible embodiments of the present application, the identification device may be a device that communicates with the monitoring apparatus through a network, and has a relatively high processing capability, and may further identify the target area human-shaped image sequence sent by the monitoring apparatus, and determine whether the target area human-shaped image sequence includes a target person.
205. The identification equipment acquires a human-shaped image sequence of the target area.
206. And the identification equipment carries out second identification on the target area human-shaped image sequence based on the pre-stored characteristics of the target person and determines whether the target area human-shaped image sequence comprises the target person. Wherein the recognition accuracy of the second recognition is greater than the recognition accuracy of the first recognition.
In a possible embodiment of the present application, the identification device may perform accurate identification on a target region human-shaped image sequence, and specifically, the identification device may perform a gait identification algorithm with any one of the following dimensional features:
and extracting and identifying the GEI characteristics, namely performing pixel-level segmentation on the human figure image sequence through a CNN model, performing superposition averaging on a human figure contour binary image segmented by the CNN model to obtain the GEI characteristics of the gait energy image, and further sending the GEI characteristics into another CNN for identification.
In some possible embodiments of the present application, keypoint feature extraction and identification may be adopted, positions of human-shaped skeleton keypoints (such as a head, a neck, a left shoulder, a left elbow, a left wrist, a right shoulder, a right elbow, a right wrist, a left hip, a right knee, a left knee, a right knee, a left ankle, and the like) in each frame of image are extracted through CNN, the position information of all frames forms a time sequence feature, the sequence is sent to a Recurrent Neural Network (RNN) to further calculate to obtain a keypoint identity feature vector, and then matching and identification are performed by calculating a similarity between the feature vector and a stored preset keypoint identity feature vector.
In some possible embodiments of the present application, optical flow feature extraction recognition may be adopted to calculate optical flow information of an object between adjacent frames through a correspondence between pixels of a previous frame and pixels of a current frame, where an optical flow is an instant speed of pixel motion and represents motion information of a pixel in an image sequence in a time domain. In some possible embodiments of the present application, a CNN may be used to directly estimate an optical flow information graph between frames from an image sequence, and then the optical flow information graph is sent to another CNN to calculate an optical flow identity feature vector, so as to perform matching identification by calculating a similarity between the feature vector and a stored preset optical flow identity feature vector.
In order to improve the identification precision of the human-shaped image sequence of the target region, in some possible embodiments of the present application, a gait identification mode in which multiple identification methods are fused may be adopted for identification. Specifically, features of three dimensions, namely human-shaped features (GEI features), key point features (such as skeleton key points and the like) and motion features (optical flow features), can be fused for gait recognition.
In particular embodiments, the method of fusion may comprise two modes. One way is that: and only performing fusion on the recognition result level, namely executing gait recognition algorithms aiming at various different dimensional characteristics respectively, and performing weighted average on the similarity corresponding to each characteristic. The other mode is as follows: the features of the three dimensions share the same convolutional neural network to extract common low-level features in the feature extraction stage, and the extracted common low-level features are further respectively input into three independent feature extraction modules to respectively obtain a human-shaped profile graph, key point positions and optical flow features. And then, respectively executing gait recognition algorithms aiming at a plurality of different dimensional characteristics, and carrying out weighted average on the similarity corresponding to each characteristic. Through the common low-level feature extraction module, gait features of all dimensions are not completely independent, correlation and constraint among the features can be utilized (for example, the known humanoid contour can help to accurately position key points and vice versa), the effect of extracting each feature is improved, and accurate identification is facilitated. After the convolutional neural network sharing network outputs the above characteristics, the following method is further adopted for identification: inputting the GEI image into a second CNN, and calculating the similarity s1 between the GEI image and a preset GEI characteristic image; inputting a time sequence formed by multi-frame key point position information into RNN, calculating to obtain a key point identity characteristic vector and the similarity s2 between the characteristic vector and a preset key point identity characteristic vector: and inputting the optical flow graph into the third CNN, and calculating to obtain an optical flow identity feature vector, namely the similarity s3 between the feature vector and a preset optical flow identity feature vector. And the final similarity S is obtained by weighted average of S1, S2 and S3, the weighted values of S1, S2 and S3 are w1, w2 and w3 respectively, and S is w 1S 1+ w 2S 2+ w 3S 3, wherein (w1+ w2+ w3 is 1).
In some possible embodiments of the present application, when the recognition device determines that the target person is included in the target area human-shaped image sequence, the recognition device may make a decision, as shown in fig. 2A, and may further include a step 207 of sending a tracking instruction to the monitoring apparatus, instructing the monitoring apparatus to adjust a monitoring angle, for example, instructing the monitoring apparatus to rotate clockwise by 10 degrees per second. 208. And the monitoring device acquires a tracking instruction sent by the identification equipment. 209. The monitoring device adjusts the monitoring angle according to the tracking instruction, such as clockwise rotation of 10 degrees per second.
In some possible embodiments of the present application, the identification device may further issue a monitoring instruction to other related monitoring apparatuses, such as extracting a motion feature including a moving direction of the target person from the target area human-shaped image sequence, and predicting an activity area of the target person according to the motion feature; and sending a monitoring instruction comprising the characteristics of the target person to a monitoring device in the activity area to perform joint monitoring on the target person.
In some possible embodiments of the application, if the similarity of the recognition result is greater than the second threshold, the recognition device may issue a target area human figure graph containing instant information such as the target person dressing attribute and the human face characteristics of the target person to a monitoring device in an adjacent area, and link with the monitoring device to perform key monitoring, tracking and recognizing the target person. (for example, a camera with human body attribute recognition capability is used for carrying out recognition and tracking based on attributes, and a camera with human face recognition capability is used for carrying out face snapshot recognition and tracking on a target person, etc.).
In some possible embodiments of the present application, when the recognition device determines that the target person is included in the target area human-shaped image sequence, the recognition device may further extract appearance characteristics of the target person from the target area human-shaped image sequence, where the appearance characteristics of the target person may include: clothing characteristics, accessory characteristics, age information, gender information, or hair style characteristics, etc. The identification device may send the appearance feature to the monitoring apparatus for the monitoring apparatus to identify the monitored video.
Referring to fig. 3A, fig. 3A is a schematic structural diagram of a monitoring device according to an embodiment of the present disclosure. The monitoring device 300 includes:
a first identifying unit 301, configured to perform first identification on a monitored video to obtain a target video frame sequence, where the target video frame sequence includes multiple target video frames;
a first processing unit 302, configured to determine a target region humanoid image sequence according to the target video frame sequence, where the target region humanoid image sequence includes a region that may include a target person in at least one target video frame of the multiple target video frames;
the first uploading unit 303 is configured to upload a target region human-shaped image sequence to a recognition device, where the target region human-shaped image sequence is used for the recognition device to determine whether the target person is included in the target region human-shaped image sequence through second recognition, and the recognition accuracy of the second recognition is greater than the recognition accuracy of the first recognition.
In some possible embodiments of the present application, the first identifying unit 301 is specifically configured to perform the first identification on each video frame in the surveillance video, so as to obtain a probability that each video frame in the surveillance video includes the target person; and composing the target video frame sequence by the video frames with the corresponding probability larger than a first preset threshold value.
In some possible embodiments of the present application, the first identifying unit 301 is specifically configured to screen out video frames including human shapes in the monitored video, so as to obtain a first video frame sequence; extracting a human body contour image from each video frame in the first video frame sequence to obtain a gait energy map; performing first identification on the gait energy map to obtain the similarity between the gait energy map and a preset gait energy map of the target person; when the similarity is greater than a second threshold, determining the first sequence of video frames as the sequence of target video frames.
In some possible embodiments of the present application, as shown in fig. 3B, the monitoring apparatus 300 may further include:
a first obtaining unit 304, configured to obtain appearance characteristics of the target person identified by the identification device; the appearance characteristic of the target person is the characteristic of the target person extracted from the target region human-shaped image sequence when the identification device identifies that the target person is included in the target region human-shaped image sequence;
the first identifying unit 301 is configured to, when a target video frame sequence is obtained by performing first identification on a monitored video, specifically, perform first identification on the monitored video by using appearance characteristics of the target person.
In some possible embodiments of the present application, as shown in fig. 3C, the first obtaining unit 304 is further configured to obtain a tracking instruction sent by the identification device.
The monitoring apparatus 300 may further include: a first adjusting unit 305, configured to adjust a monitoring angle of the monitoring device according to the tracking instruction.
Referring to fig. 4A, fig. 4A is a schematic structural diagram of an identification device according to an embodiment of the present application. The recognition device 400 may include:
a second obtaining unit 401, configured to obtain a target area human-shaped image sequence uploaded by the monitoring terminal; the target area humanoid image sequence is determined by a target video frame sequence, the target video frame sequence comprises a plurality of target video frames, the target video frame sequence is obtained by performing first identification on a monitoring video by the monitoring terminal, and the target area humanoid image sequence comprises an area which possibly comprises a target person in at least one target video frame in the plurality of target video frames;
a second identifying unit 402, configured to perform second identification on the target region human-shaped image sequence based on a pre-stored feature of the target person, and determine whether the target region human-shaped image sequence includes the target person, where an identification accuracy of the second identification is greater than an identification accuracy of the first identification.
In some possible embodiments of the present application, as shown in fig. 4B, the identification device 400 may further include:
a first sending unit 403, configured to send a tracking instruction to the monitoring terminal, where the tracking instruction is used to instruct the monitoring terminal to adjust a monitoring angle.
In some possible embodiments of the present application, the pre-stored characteristics of the target person may include one or more of the following characteristics: gait energy map of the target person, face features of the target person, human shape features of the target person, skeleton features of the target person, optical flow features of the target person, and the like.
In some possible embodiments of the present application, as shown in fig. 4C, the identification device 400 may further include:
a second processing unit 404, configured to, when the second identification unit determines that the target person is included in the target region human-shaped image sequence, extract a motion feature of the target person from the target region human-shaped image sequence, where the motion feature includes a moving direction; predicting the activity area of the target person according to the motion characteristics;
a first sending unit 403, configured to send a monitoring instruction including characteristics of the target person to a monitoring apparatus in the activity area.
In some possible embodiments of the present application, as shown in fig. 4C, the identification device 400 may further include:
a second processing unit 404, configured to, when the second identifying unit 402 determines that the target person is included in the target region human-shaped image sequence, extract appearance characteristics of the target person from the target region human-shaped image sequence.
A first sending unit 403, configured to send the appearance characteristics of the target person to a monitoring terminal, where the appearance characteristics of the target person are used for the monitoring terminal to perform first identification on the monitored video.
In some possible embodiments of the present application, the second identification unit 402 is configured to extract two or more pre-stored features of the target person based on the pre-stored features of the target person when performing the second identification on the target region figure image sequence based on the pre-stored features of the target person, and identify the target region figure image sequence based on the extracted two or more pre-stored features of the target person.
An embodiment of the present application further provides a computer-readable storage medium, on which a program is stored, and when the program runs, the monitoring method executed by any of the foregoing embodiments of the monitoring apparatus is implemented.
The embodiment of the present application further provides a computer-readable storage medium, on which a program is stored, and when the program runs, the method for recognizing the device performed in any of the foregoing embodiments is implemented.
An embodiment of the present application further provides a monitoring apparatus, as shown in fig. 5, the monitoring apparatus 500 includes: a camera 501, a communication unit 502, a processor 503, a memory 504 and a bus 505; the camera 501 is used for acquiring a monitoring video; a communication unit 502 for communicating with an identification device; the camera 501, the communication unit 502, the processor 503 and the memory 504 are connected through the bus 505 and complete mutual communication; the memory 504 stores executable program code; the processor 503 runs a program corresponding to the executable program code by reading the executable program code stored in the memory 504, so as to execute the monitoring method executed by the monitoring apparatus in any of the foregoing embodiments.
An embodiment of the present application further provides an identification device, as shown in fig. 6, an identification device 600 includes: a communication unit 601, a memory 602, a processor 603, and a bus 604; the communication unit 601 is used for communicating with the monitoring device; the memory 602 stores executable program code; the processor 603 runs a program corresponding to the executable program code by reading the executable program code stored in the memory 602 for performing the identification method performed by the identification device in any of the previous embodiments.
The embodiment of the present application further provides a monitoring system, which includes a monitoring device and an identification device, where the monitoring device may be the monitoring device described in any of the foregoing embodiments, and the identification device may be the identification device described in any of the foregoing embodiments.
Those skilled in the art will recognize that, in one or more of the examples described above, the functions described in connection with the embodiments of the invention may be implemented in hardware, software, firmware, or any combination thereof. When implemented in software, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.
The above embodiments are only intended to be illustrative of the embodiments of the present invention, and should not be construed as limiting the scope of the embodiments of the present invention, and any modifications, equivalent substitutions, improvements and the like made on the basis of the technical solutions of the embodiments of the present invention should be included in the scope of the embodiments of the present invention.
Claims (25)
1. A method of monitoring, the method comprising:
performing first identification processing on a monitoring video, and screening out video frames including human shapes in the monitoring video to obtain a first video frame sequence;
extracting a human-shaped contour image from each video frame in the first video frame sequence to obtain a gait energy map;
performing first identification on the gait energy map to obtain the similarity between the gait energy map and a preset gait energy map of the target person;
when the similarity is greater than a second threshold, determining the first video frame sequence as the target video frame sequence, wherein the target video frame sequence comprises a plurality of target video frames;
determining a target region humanoid image sequence according to the target video frame sequence, wherein the target region humanoid image sequence comprises a region which is possibly provided with a target person in at least one target video frame in the plurality of target video frames;
uploading the target area human-shaped image sequence to a recognition device, wherein the target area human-shaped image sequence is used for determining whether the target person is included in the target area human-shaped image sequence or not through second recognition by the recognition device, and the recognition accuracy of the second recognition is greater than that of the first recognition.
2. The method of claim 1, wherein the first identifying a surveillance video to obtain a sequence of target video frames comprises:
performing the first identification on each video frame in the monitored video to obtain the probability that each video frame in the monitored video comprises the target person;
and composing the target video frame sequence by the video frames with the corresponding probability larger than a first preset threshold value.
3. The method of claim 1, further comprising:
acquiring the appearance characteristics of the target person identified by the identification equipment;
the first identification of the monitoring video to obtain a target video frame sequence comprises the following steps: and performing the first identification on the monitoring video according to the appearance characteristics to obtain the target video frame sequence.
4. The method according to any one of claims 1 to 3, further comprising:
acquiring a tracking instruction sent by the identification equipment;
and adjusting the monitoring angle according to the tracking instruction.
5. An identification method, characterized in that the method comprises:
acquiring a target area human-shaped image sequence uploaded by a monitoring device; the target region humanoid image sequence is determined by a target video frame sequence, the target video frame sequence comprises a plurality of target video frames, the target video frame sequence is obtained by performing first identification on a monitoring video through the monitoring device, and the target region humanoid image sequence comprises a region which possibly comprises a target person in at least one target video frame in the plurality of target video frames;
and performing second recognition on the target region human-shaped image sequence based on the pre-stored characteristics of the target person, and determining whether the target region human-shaped image sequence comprises the target person, wherein the recognition accuracy of the second recognition is greater than that of the first recognition.
6. The method of claim 5, further comprising:
and sending a tracking instruction to the monitoring device, wherein the tracking instruction is used for instructing the monitoring device to adjust a monitoring angle.
7. The method of claim 5,
the pre-stored characteristics of the target person include one or more of the following characteristics: a gait energy map of the target person, facial features of the target person, anthropomorphic features of the target person, skeletal features of the target person, and optical flow features of the target person.
8. The method of claim 5, wherein upon determining that the target person is included in the sequence of target area humanoid images, the method further comprises:
extracting motion characteristics of the target person from the target area humanoid image sequence, wherein the motion characteristics comprise a moving direction;
predicting the activity area of the target person according to the motion characteristics;
sending a monitoring instruction including characteristics of the target person to a monitoring device in the activity area.
9. The method of claim 5, further comprising: when it is determined that the target person is included in the target area humanoid image sequence, the method further includes:
extracting appearance characteristics of the target person from the target area humanoid image sequence;
and sending the appearance characteristics of the target person to the monitoring device, wherein the appearance characteristics of the target person are used for the monitoring device to perform the first identification on the monitored video.
10. The method according to any one of claims 5 to 9, wherein when the pre-stored characteristics of the target person include two or more characteristics, the second recognition of the target area human-shaped image sequence based on the pre-stored characteristics of the target person comprises:
and extracting two or more pre-stored characteristics of the target figure by utilizing the target area figure image sequence, and performing the second identification on the target area figure image sequence based on the extracted two or more pre-stored characteristics of the target figure.
11. A monitoring device, characterized in that the monitoring device comprises:
the first identification unit is used for carrying out first identification processing on a monitored video, screening out video frames including human shapes in the monitored video and obtaining a first video frame sequence; extracting a human-shaped contour image from each video frame in the first video frame sequence to obtain a gait energy map; performing first identification on the gait energy map to obtain the similarity between the gait energy map and a preset gait energy map of the target person; when the similarity is greater than a second threshold, determining the first video frame sequence as the target video frame sequence, wherein the target video frame sequence comprises a plurality of target video frames;
a first processing unit, configured to determine a target region humanoid image sequence according to the target video frame sequence, where the target region humanoid image sequence includes a region that may include a target person in at least one target video frame of the multiple target video frames;
the first uploading unit is used for uploading the target area human-shaped image sequence to the recognition equipment, the target area human-shaped image sequence is used for determining whether the target area human-shaped image sequence comprises the target person or not through second recognition by the recognition equipment, and the recognition accuracy of the second recognition is greater than that of the first recognition.
12. The monitoring device of claim 11,
the first identification unit is specifically configured to perform the first identification on each video frame in the surveillance video to obtain a probability that each video frame in the surveillance video includes the target person; and composing the target video frame sequence by the video frames with the corresponding probability larger than a first preset threshold value.
13. The monitoring device of claim 11, further comprising:
a first obtaining unit, configured to obtain appearance characteristics of the target person identified by the identification device;
the first identification unit is used for performing first identification on the monitored video to obtain a target video frame sequence, and is specifically used for performing first identification on the monitored video by using the appearance characteristics of the target person.
14. The monitoring device of claim 13,
the first obtaining unit is further configured to obtain a tracking instruction sent by the identification device;
the monitoring device further comprises:
and the first adjusting unit is used for adjusting the monitored angle according to the tracking instruction.
15. An identification device, characterized in that the identification device comprises:
the second acquisition unit is used for acquiring a target area human-shaped image sequence uploaded by the monitoring device; the target region humanoid image sequence is determined by a target video frame sequence, the target video frame sequence comprises a plurality of target video frames, the target video frame sequence is obtained by performing first identification on a monitoring video through the monitoring device, and the target region humanoid image sequence comprises a region which possibly comprises a target person in at least one target video frame in the plurality of target video frames;
and the second identification unit is used for carrying out second identification on the target area human-shaped image sequence based on the pre-stored characteristics of the target person and determining whether the target area human-shaped image sequence comprises the target person, and the identification accuracy of the second identification is greater than that of the first identification.
16. The identification device of claim 15, further comprising:
the monitoring device comprises a first sending unit, a second sending unit and a control unit, wherein the first sending unit is used for sending a tracking instruction to the monitoring device, and the tracking instruction is used for indicating the monitoring device to adjust a monitoring angle.
17. The identification device of claim 15,
the pre-stored characteristics of the target person include one or more of the following characteristics: a gait energy map of the target person, facial features of the target person, anthropomorphic features of the target person, skeletal features of the target person, and optical flow features of the target person.
18. The identification device of claim 15, further comprising:
the second processing unit is used for extracting the motion characteristics of the target person from the target area human-shaped image sequence when the second identification unit determines that the target person is included in the target area human-shaped image sequence, and the motion characteristics comprise the moving direction; predicting the activity area of the target person according to the motion characteristics;
a first sending unit, configured to send a monitoring instruction including characteristics of the target person to a monitoring apparatus in the activity area.
19. The identification device according to claim 15, characterized in that the identification device comprises further comprising:
the second processing unit is used for extracting the appearance characteristics of the target person from the target area human-shaped image sequence when the second identification unit determines that the target person is included in the target area human-shaped image sequence;
and the first sending unit is used for sending the appearance characteristics of the target person to the monitoring device, and the appearance characteristics of the target person are used for the monitoring device to perform first identification on the monitored video.
20. Identification device according to any of claims 15-19,
the second identification unit is specifically configured to extract two or more pre-stored features of the target person from the target region figure image sequence and identify the target region figure image sequence based on the extracted two or more pre-stored features of the target person, when the two or more pre-stored features of the target person are used for performing second identification on the target region figure image sequence based on the pre-stored features of the target person.
21. A computer-readable storage medium, characterized in that the storage medium has stored thereon a program which, when executed, implements the monitoring method according to any one of claims 1-4.
22. A computer-readable storage medium, characterized in that the storage medium has stored thereon a program which, when executed, carries out the identification method according to any one of claims 5-10.
23. A monitoring device, comprising: the device comprises a camera, a communication unit, a processor, a memory and a bus; wherein the content of the first and second substances,
the camera is used for acquiring a monitoring video;
the communication unit is used for communicating with the identification equipment;
the camera, the communication unit, the processor and the memory are connected through the bus and complete mutual communication;
the memory stores executable program code;
the processor executes a program corresponding to the executable program code by reading the executable program code stored in the memory for executing the monitoring method according to any one of claims 1 to 4.
24. An identification device, comprising: a communication unit, a memory, a processor and a bus; wherein the content of the first and second substances,
the communication unit is used for communicating with the monitoring device;
the memory stores executable program code;
the processor executes a program corresponding to the executable program code by reading the executable program code stored in the memory for performing the identification method according to any one of claims 5 to 10.
25. A monitoring system comprising a monitoring apparatus according to claim 23 and an identification device according to claim 24.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111170226.2A CN114049681A (en) | 2018-06-19 | 2018-06-19 | Monitoring method, identification method, related device and system |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111170226.2A CN114049681A (en) | 2018-06-19 | 2018-06-19 | Monitoring method, identification method, related device and system |
| CN201810630204.1A CN108600707A (en) | 2018-06-19 | 2018-06-19 | A kind of monitoring method, recognition methods, relevant apparatus and system |
Related Parent Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810630204.1A Division CN108600707A (en) | 2018-06-19 | 2018-06-19 | A kind of monitoring method, recognition methods, relevant apparatus and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114049681A true CN114049681A (en) | 2022-02-15 |
Family
ID=63628664
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810630204.1A Pending CN108600707A (en) | 2018-06-19 | 2018-06-19 | A kind of monitoring method, recognition methods, relevant apparatus and system |
| CN202111170226.2A Pending CN114049681A (en) | 2018-06-19 | 2018-06-19 | Monitoring method, identification method, related device and system |
Family Applications Before (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810630204.1A Pending CN108600707A (en) | 2018-06-19 | 2018-06-19 | A kind of monitoring method, recognition methods, relevant apparatus and system |
Country Status (2)
| Country | Link |
|---|---|
| CN (2) | CN108600707A (en) |
| WO (1) | WO2019242330A1 (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108600707A (en) * | 2018-06-19 | 2018-09-28 | 华为技术有限公司 | A kind of monitoring method, recognition methods, relevant apparatus and system |
| CN111079477A (en) * | 2018-10-19 | 2020-04-28 | 北京奇虎科技有限公司 | Monitoring analysis method and monitoring analysis system |
| CN109670451A (en) * | 2018-12-20 | 2019-04-23 | 天津天地伟业信息系统集成有限公司 | Automatic face recognition tracking |
| CN110163146A (en) * | 2019-05-21 | 2019-08-23 | 银河水滴科技（北京）有限公司 | A kind of monitoring method and device based on characteristics of human body |
| CN110062212A (en) * | 2019-05-22 | 2019-07-26 | 广东慧讯智慧科技有限公司 | Video monitoring method, device, system, equipment and computer readable storage medium |
| CN110287890A (en) * | 2019-06-26 | 2019-09-27 | 银河水滴科技（北京）有限公司 | A kind of recognition methods and device based on gait feature and pedestrian's weight identification feature |
| CN111291758B (en) * | 2020-02-17 | 2023-08-04 | 北京百度网讯科技有限公司 | Method and device for recognizing seal characters |
| CN111881726B (en) * | 2020-06-15 | 2022-11-25 | 马上消费金融股份有限公司 | Living body detection method and device and storage medium |
| CN111985321A (en) * | 2020-07-14 | 2020-11-24 | 浙江大华技术股份有限公司 | Target object detection method, electronic device, and storage medium |
| CN112218046B (en) * | 2020-09-27 | 2023-10-24 | 杭州海康威视系统技术有限公司 | Object monitoring method and device |
| CN112784740B (en) * | 2021-01-21 | 2022-10-04 | 上海市公安局刑事侦查总队 | Gait data acquisition and labeling method and application |
| CN113221815A (en) * | 2021-05-25 | 2021-08-06 | 北京无垠创新科技有限责任公司 | Gait identification method based on automatic detection technology of skeletal key points |
| CN113364911B (en) * | 2021-06-11 | 2023-03-07 | 上海兴容信息技术有限公司 | Detection method and system for preset terminal |
| CN113673342A (en) * | 2021-07-19 | 2021-11-19 | 浙江大华技术股份有限公司 | Behavior detection method, electronic device, and storage medium |
| CN113705510A (en) * | 2021-09-02 | 2021-11-26 | 广州市奥威亚电子科技有限公司 | Target identification tracking method, device, equipment and storage medium |
| CN117372968B (en) * | 2023-12-07 | 2024-03-22 | 深圳市智慧城市通信有限公司 | Electric vehicle home-entering monitoring method based on Internet of things |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101739685B (en) * | 2009-02-11 | 2012-04-18 | 北京智安邦科技有限公司 | Moving object classification method and system thereof |
| CN101635031B (en) * | 2009-08-27 | 2011-08-31 | 上海交通大学 | Method for extracting and identifying small sample character contour feature |
| JP6335551B2 (en) * | 2014-03-03 | 2018-05-30 | キヤノン株式会社 | Image forming apparatus, monitoring system, log management method, and computer program |
| CN106182024A (en) * | 2016-07-11 | 2016-12-07 | 天津艾思科尔科技有限公司 | A kind of available image carries out the household service robot of gait detection |
| CN107071344A (en) * | 2017-01-22 | 2017-08-18 | 深圳英飞拓科技股份有限公司 | A kind of large-scale distributed monitor video data processing method and device |
| CN107403143A (en) * | 2017-07-06 | 2017-11-28 | 广东小天才科技有限公司 | A kind of gait recognition method and electronic equipment |
| CN108600707A (en) * | 2018-06-19 | 2018-09-28 | 华为技术有限公司 | A kind of monitoring method, recognition methods, relevant apparatus and system |
-
2018
- 2018-06-19 CN CN201810630204.1A patent/CN108600707A/en active Pending
- 2018-06-19 CN CN202111170226.2A patent/CN114049681A/en active Pending
-
2019
- 2019-03-07 WO PCT/CN2019/077293 patent/WO2019242330A1/en active Application Filing
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019242330A1 (en) | 2019-12-26 |
| CN108600707A (en) | 2018-09-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114049681A (en) | Monitoring method, identification method, related device and system | |
| CN108229369B (en) | Image shooting method and device, storage medium and electronic equipment | |
| WO2019128507A1 (en) | Image processing method and apparatus, storage medium and electronic device | |
| WO2019128508A1 (en) | Method and apparatus for processing image, storage medium, and electronic device | |
| WO2019137131A1 (en) | Image processing method, apparatus, storage medium, and electronic device | |
| TW201911130A (en) | Method and device for remake image recognition | |
| CN111062263B (en) | Method, apparatus, computer apparatus and storage medium for hand gesture estimation | |
| Kalas | Real time face detection and tracking using OpenCV | |
| CN108875507B (en) | Pedestrian tracking method, apparatus, system, and computer-readable storage medium | |
| CN111626113A (en) | Facial expression recognition method and device based on facial action unit | |
| CN111339831A (en) | Lighting lamp control method and system | |
| CN109002776B (en) | Face recognition method, system, computer device and computer-readable storage medium | |
| CN110807391A (en) | Human body posture instruction identification method for human-unmanned aerial vehicle interaction based on vision | |
| WO2023279799A1 (en) | Object identification method and apparatus, and electronic system | |
| CN111353429A (en) | Interest degree method and system based on eyeball turning | |
| CN110266953B (en) | Image processing method, image processing apparatus, server, and storage medium | |
| CN111914762A (en) | Gait information-based identity recognition method and device | |
| CN109214278B (en) | User instruction matching method and device, computer equipment and storage medium | |
| CN110909691A (en) | Motion detection method and device, computer readable storage medium and computer equipment | |
| Ahlvers et al. | Model-free face detection and head tracking with morphological hole mapping | |
| CN113496176A (en) | Motion recognition method and device and electronic equipment | |
| CN111507289A (en) | Video matching method, computer device and storage medium | |
| CN113421182B (en) | Three-dimensional reconstruction method, three-dimensional reconstruction device, electronic equipment and storage medium | |
| CN114387670A (en) | Gait recognition method and device based on space-time feature fusion and storage medium | |
| CN116580054A (en) | Video data processing method, device, equipment and medium |
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 |