CN104463117A - Sample collection method and system used for face recognition and based on video - Google Patents
Sample collection method and system used for face recognition and based on video Download PDFInfo
- Publication number
- CN104463117A CN104463117A CN201410720464.XA CN201410720464A CN104463117A CN 104463117 A CN104463117 A CN 104463117A CN 201410720464 A CN201410720464 A CN 201410720464A CN 104463117 A CN104463117 A CN 104463117A
- Authority
- CN
- China
- Prior art keywords
- face
- characteristic
- pixel region
- picture frame
- recognition
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
Abstract
The invention provides a sample collection method and system used for face recognition and based on a video. The sample collection method used for face recognition and based on the video comprises the steps that an image region, containing movement information, in an image frame of the video is obtained firstly, the image region is detected so that a face pixel region can be obtained, the face pixel region in the image frame containing the face is tracked, so that a historical image frame of the same face is obtained, face characteristic extraction is conducted on the high-quality historical image frame of the same face, so that face characteristics required for face recognition are obtained, and finally, the face characteristics are stored. According to the sample collection method and system used for face recognition and based on the video, manual operation of sample collection workers is not needed in the whole sample collection process for face recognition, the collection speed is high, and the collection efficiency is high.
Description
Technical field
The present invention relates to technical field of face recognition, particularly a kind of recognition of face sample collection method and system based on video mode.
Background technology
The study hotspot of artificial intelligence field is not only in recognition of face, and at public safety field, its research also has important realistic meaning.Face recognition technology comprises facial image sample collection, sample image pre-service, sorter training and these sport technique segments of specimen discerning, and sample collection work is the foundation stone of recognition of face, and its meaning is self-evident.
Recognition of face sample collection work was all the mode based on image in the past.Such as, first allow volunteer is orderly is successively in certain shooting point; Then, sample collection worker repeatedly takes volunteer according to differences such as attitude, illumination, expressions; Finally, sample collection worker carries out Screening Treatment according to taking the image obtained, and sets up recognition of face Sample Storehouse.Although this mode based on image is directly perceived, but still there are 3 deficiencies: the first, allow a large amount of volunteers shooting point that is in orderly successively repeatedly take pictures, this way extremely takes time and effort; The second, recognition of face has strict requirement to face sample in resolution, and therefore, can the image of shooting reach criterion of identification, and sample collection worker need be wanted to check one by one; 3rd, it is a long process that sample arranges, and sample collection worker produces work fatigue unavoidably when arranging, and therefore, the face information of warehouse-in also exists the risk of mistake.
Also some face recognition sample collecting works are had based on video mode in prior art.Such as, with CCTV camera in a certain fixed point capture video, sample collection worker intercepts the picture frame containing face in video on backstage, then identifies according to attitude, illumination, expression etc. the picture frame containing face, the picture frame met is put into recognition of face Sample Storehouse.This sample collection technology based on video mode, volunteer's shooting point that is in orderly successively is not needed repeatedly to take, but still need sample collection worker manually to choose picture frame containing face, and whether artificial this picture frame of examination reaches criterion of identification, then carry out sample arrangement.Whole process wastes time and energy, and efficiency is low.
Summary of the invention
For this reason, technical matters to be solved by this invention to be in prior art based on slow, the consuming time length of recognition of face sample collection method speed of video and efficiency is low, thus proposes a kind of recognition of face sample collection method and system based on video mode.
For solving the problems of the technologies described above, the invention provides following technical scheme:
Based on a recognition of face sample collection method for video mode, comprise the steps:
Obtain the image-region in the picture frame of video with movable information;
Detect described image-region to obtain face pixel region;
Follow the tracks of containing face pixel region in the picture frame of face, obtain the historigram picture frame of same face;
Face characteristic extraction is carried out to the historigram picture frame of the high same face of quality, obtains the face characteristic needed for recognition of face;
Store described face characteristic.
The above-mentioned recognition of face sample collection method based on video mode, the image-region step in the picture frame of described acquisition video with movable information comprises:
The initial image frame in described video is utilized to set up background model;
Foreground model is set up to current image frame;
Calculate the difference of described background model and described foreground model;
The difference obtained between foreground model and background model is greater than the pixel region of predetermined threshold value, as the image-region with movable information.
The above-mentioned recognition of face sample collection method based on video mode, the described image-region of described detection comprises to obtain face pixel region step:
Image pyramid is set up to current image frame;
Be loaded into each attitude faceform of off-line training;
Calculate the feature on each tomographic image pyramid;
Calculate the response of described feature on each attitude faceform of off-line training;
Obtain the picture frame that maximum response reaches Face datection threshold value, extract the attitude information of face pixel region and face in current image frame.
The above-mentioned recognition of face sample collection method based on video mode, describedly tracking step is carried out to face pixel region in the described picture frame containing face comprise:
Set up face tracking model and face Track Initiation node;
Calculate textural characteristics and the response of edge feature on face tracking model of face pixel region in current image frame;
Face pixel region coordinate in current image frame corresponding to peak response is added the afterbody node of face track;
If when the face pixel region of current image frame arrives the border of picture frame, terminate to follow the tracks of, obtain the historigram picture frame from face Track Initiation node to the same face of face track afterbody node.
The above-mentioned recognition of face sample collection method based on video mode, the historigram picture frame of the described same face high to quality carries out face characteristic extraction step and comprises:
The picture frame high to quality carries out illumination pretreatment;
Calculate the face characteristic of face pixel region in the picture frame after illumination pretreatment;
Principle components analysis and independent component analysis are carried out to described face characteristic;
Obtain the face characteristic needed for recognition of face.
The above-mentioned recognition of face sample collection method based on video mode, the described face characteristic step of described storage comprises:
Each face characteristic in described face characteristic and face Sample Storehouse is compared;
If some face characteristics are close in described face characteristic and face Sample Storehouse, then the face pixel region of described face characteristic and correspondence is stored in this close face characteristic sample information Sample Storehouse;
If each face characteristic differs greatly in described face characteristic and face Sample Storehouse, then a newly-built face characteristic information Sample Storehouse stores the face pixel region of described face characteristic and correspondence.
Based on a recognition of face sample acquisition system for video mode, comprising:
Image-region acquisition module, for obtain video picture frame in there is the image-region of movable information;
Face detection module, for detecting described image-region to obtain face pixel region;
Face tracking module, for following the tracks of containing face pixel region in the picture frame of face, obtains the historigram picture frame of same face;
Face characteristic acquisition module, for carrying out face characteristic extraction to the historigram picture frame of the high same face of quality, obtains the face characteristic needed for recognition of face;
Face characteristic memory module, for storing described face characteristic.
The above-mentioned recognition of face sample acquisition system based on video mode, described image-region acquisition module comprises:
Background Modeling submodule, sets up background model for utilizing the initial image frame in described video;
Foreground model sets up submodule, for setting up foreground model to current image frame;
Difference calculating sub module, for calculating the difference of described background model and described foreground model;
Image-region obtains submodule, is greater than the pixel region of predetermined threshold value, as the image-region with movable information for the difference obtained between foreground model and background model.
The above-mentioned recognition of face sample acquisition system based on video mode, in described face detection module:
Image pyramid sets up submodule, for setting up image pyramid to current image frame;
Off-line model is loaded into submodule, for being loaded into each attitude faceform of off-line training;
Feature calculation submodule, for calculating the feature on each tomographic image pyramid;
Response calculating sub module, for calculating the response of described feature on each attitude faceform of off-line training;
Face pixel region obtains submodule, reaches the picture frame of Face datection threshold value for obtaining maximum response, extracts the attitude information of face pixel region and face in current image frame.
The above-mentioned recognition of face sample acquisition system based on video mode, described face tracking module comprises:
Submodule set up by face tracking model, for setting up face tracking model and face Track Initiation node;
Characteristic response calculating sub module, for calculating textural characteristics and the response of edge feature on face tracking model of face pixel region in current image frame;
Peak response process submodule, for adding the afterbody node of face track by the face pixel region coordinate in current image frame corresponding to peak response;
The historigram picture frame of same face obtains submodule, for when the face pixel region of current image frame arrives the border of picture frame, terminates to follow the tracks of, obtains the historigram picture frame from face Track Initiation node to the same face of face track afterbody node.
The above-mentioned recognition of face sample acquisition system based on video mode, described face characteristic acquisition module comprises:
Pre-service submodule, carries out illumination pretreatment for the picture frame high to quality;
Face characteristic calculating sub module, for calculating the face characteristic of face pixel region in the picture frame after illumination pretreatment;
Human face analysis submodule, for carrying out Principle components analysis and independent component analysis to described face characteristic;
Face characteristic obtains submodule, for obtaining the face characteristic needed for recognition of face.
The above-mentioned recognition of face sample acquisition system based on video mode, described face characteristic memory module comprises:
Face characteristic comparer module, for comparing each face characteristic in face characteristic described in face characteristic acquisition module and face Sample Storehouse;
Face characteristic sub module stored, if close for some face characteristics in described face characteristic and face Sample Storehouse, is then stored in this close face characteristic sample information Sample Storehouse by the face pixel region of described face characteristic and correspondence; If each face characteristic differs greatly in described face characteristic and face Sample Storehouse, then a newly-built face characteristic information Sample Storehouse stores the face pixel region of described face characteristic and correspondence.
Technique scheme of the present invention has the following advantages compared to existing technology:
(1) a kind of recognition of face sample collection method and system based on video mode of the present invention, first the image-region in the picture frame of video with movable information is obtained, then described image-region is detected to obtain face pixel region, then follow the tracks of containing face pixel region in the picture frame of face, obtain the historigram picture frame of same face, again face characteristic extraction is carried out to the historigram picture frame of the high same face of quality, obtain the face characteristic needed for recognition of face, finally store described face characteristic.The process of whole recognition of face sample collection does not need sample collection worker manual operation, and the speed of collection is fast, efficiency is high.
Accompanying drawing explanation
In order to make content of the present invention be more likely to be clearly understood, below according to a particular embodiment of the invention and by reference to the accompanying drawings, the present invention is further detailed explanation, wherein
Fig. 1 is a kind of recognition of face sample collection method flow diagram based on video mode of one embodiment of the invention;
Fig. 2 is the method flow diagram that a kind of acquisition of one embodiment of the invention has the image-region of movable information;
Fig. 3 is the method flow diagram of a kind of Face datection of one embodiment of the invention;
Fig. 4 is the method flow diagram of a kind of face tracking of one embodiment of the invention;
Fig. 5 is the method flow diagram that a kind of face characteristic of one embodiment of the invention obtains;
Fig. 6 is a kind of recognition of face sample collection information network topology diagram of one embodiment of the present of invention;
Fig. 7 is a kind of recognition of face sample collection method flow diagram of one embodiment of the invention;
Fig. 8 is a kind of recognition of face sample acquisition system block diagram based on video mode of one embodiment of the invention.
In figure, Reference numeral is expressed as: the common collection front end of 1-, 2-intelligent electronic device, 3-front-end processor, 4-first server, 5-second server, 6-first client, 7-second client.
Embodiment
Embodiment 1
The present embodiment provides a kind of recognition of face sample collection method based on video mode, as shown in Figure 1, comprises the steps:
S1: obtain the image-region in the picture frame of video with movable information, as shown in Figure 2, comprising:
First utilize the initial image frame of video to set up the background model of each pixel or pixel region, initial image frame can choose the 1st frame or front 3 frames.
Foreground model is set up to pixel each in current image frame or pixel region.
Calculate the difference between the background model of each pixel or pixel region and foreground model.
Determine a predetermined threshold value 20*20 pixel, if the connected region producing the pixel region of described difference is less than described predetermined threshold value, illustrate that current frame image does not produce movable information, face can not be there is, deletion can not produce the picture frame of face, upgrade background model, if the connected region producing the pixel region of described difference is more than or equal to described predetermined threshold value, illustrate that current frame image produces movable information, may face be there is, retain present frame and obtain described pixel region as the image-region with movable information.
S2: detect described image-region to obtain face pixel region, as shown in Figure 3, comprising:
Image pyramid is set up to current frame image, convergent-divergent is carried out and synthetic image pyramid to the image-region with movable information obtained in step S1.
Being loaded into each attitude faceform of off-line training, is 20*20 pixel for the face pixel region of Face datection in described multi-pose Face model.Each attitude faceform of described off-line training is that off-line training is good in advance, just can use as long as be loaded in the method.The off-line training process of each attitude faceform is: first the image containing real human face gathered is stored to the positive Sample Storehouse of face; The image not containing face gathered is stored to face negative example base; By sample collection, worker manually demarcates, determine face in positive Sample Storehouse the accurate coordinates gathered in image and attitude towards; From collection image, extract the less face pixel region of background interference according to face coordinate information, and calculate the feature of described pixel region, described in be characterized as Haar feature; Described feature is integrated, obtains each attitude faceform of off-line training.
Calculate the feature on each tomographic image pyramid, described in be characterized as Haar feature.
Calculate the response of described feature on each attitude faceform of off-line training, by the feature on each tomographic image pyramid is carried out filtering operation thus the value that meets with a response on each attitude faceform of off-line training.
The described response of comprehensive analysis, obtains maximum response, if maximum response is less than Face datection threshold value, then exports non-face mark, deletes current image frame; If maximum response is more than or equal to Face datection threshold value, then exports face mark, extract the attitude information of face pixel region and face in current image frame.
S3: follow the tracks of containing face pixel region in the picture frame of face, obtains the historigram picture frame of same face, as shown in Figure 4, comprising:
Set up face tracking model and face Track Initiation node, first judge in the current image frame obtained in step s 2, whether face pixel region has had face tracking model: if do not had, then calculate the texture of human face region and edge feature and form face tracking model, and face pixel region coordinate is added face Track Initiation node, described face tracking model can be made up of gang's Haar wave filter; If fruit has, then continue next step.
Calculate textural characteristics and the response of edge feature on face tracking model of face pixel region in current image frame;
Face pixel region coordinate in current image frame corresponding to peak response is added the afterbody node of face track;
If the face pixel region of current image frame has arrived the border of picture frame, terminate to follow the tracks of, obtained the historigram picture frame from face Track Initiation node to the same face of face track afterbody node; Otherwise, turn back to and describedly set up face tracking model and face Track Initiation node step.
S4: carry out face characteristic extraction to the historigram picture frame of the high same face of quality, obtains the face characteristic needed for recognition of face, as shown in Figure 5, comprising:
The picture frame high to quality carries out illumination pretreatment.The acquisition methods of the picture frame that described quality is high is carry out face evaluation process to the historigram picture frame of the same face obtained in step S3: carry out comprehensive evaluation according to the resolution of face, human face posture, sharpness and the face pixel region of symmetry to each picture frame in the historigram picture frame of same face and provide quality score, sort according to each picture frame in the historigram picture frame of the same face of described quality score, obtain described quality score and reach the picture frame of setting threshold value as the high picture frame of quality.Face pixel region to be zoomed to the face resolution of restriction by the picture frame high to quality after carrying out illumination pretreatment, generally choose 96*96 pixel.
Calculate the face characteristic of face pixel region in the picture frame after illumination pretreatment.Set 40 Gabor cores in 5 yardsticks, 8 directions, check face pixel region in the picture frame after photo-irradiation treatment by described Gabor and do convolution, then carry out piecemeal process acquisition face characteristic.
Principle components analysis and independent component analysis are carried out to described face characteristic.Reduce the Linear Redundancy degree between described face characteristic by Principle components analysis, promote the discrimination between described face characteristic by independent component analysis, realize the dimensionality reduction to described face characteristic.
Obtain the face characteristic needed for recognition of face.
S5: store described face characteristic, comprising:
Each face characteristic in the described face characteristic obtained in step S4 and face Sample Storehouse is compared, Euclidean distance can be adopted when comparison to compare.
If some face characteristics are close in described face characteristic and face Sample Storehouse, then the face pixel region of described face characteristic and correspondence is stored in this close face characteristic sample information Sample Storehouse;
If each face characteristic differs greatly in described face characteristic and face Sample Storehouse, then a newly-built face characteristic information Sample Storehouse stores the face pixel region of described face characteristic and correspondence.
Before step S1, also comprise step:
S0: the video obtaining shooting.Minimum face imaging resolution determination camera pedestal required for the inside and outside parameter of video camera and recognition of face sets up an office and takes working point, then obtains the video of video camera shooting.Minimum face imaging resolution can limit face two eye pupil of recognition of face apart from being not less than 50*50 pixel, or the face pixel region of recognition of face is not less than 96*96 pixel, chooses face pixel region in the present embodiment and is not less than 96*96 pixel.General, video camera erection is highly overhead 2.5-4 rice, and monitoring width is about 3 meters, and the camera angle of pitch is within 15 degree, and the focal length of camera lens is 25-50mm.
A kind of recognition of face sample collection method based on video mode described in the present embodiment, first the image-region in the picture frame of video with movable information is obtained, then described image-region is detected to obtain face pixel region, then follow the tracks of containing face pixel region in the picture frame of face, obtain the historigram picture frame of same face, again face characteristic extraction is carried out to the historigram picture frame of the high same face of quality, obtain the face characteristic needed for recognition of face, finally store described face characteristic.The process of whole recognition of face sample collection does not need sample collection worker manual operation, and the speed of collection is fast, efficiency is high.
Embodiment 2
The present embodiment provides a kind of recognition of face sample collection method based on video mode.Recognition of face sample collection information network topological diagram as shown in Figure 6, mainly contains according to function and gathers front end, back-end server and customer account management end three parts.Gather front end and have common collection front end 1 and intelligent electronic device 2 two kinds of forms, front-end processor 3 exports head end video file, video image after the output encoder of common collection front end 1, the output of intelligent electronic device 2 comprises the part face information that intelligent front end extracts, and specifically refers to the information through forms such as concentrated frame of video, face snap area image or face characteristic data.Back-end server comprises first server 4 and second server 5, the video of front-end collection is carried out face characteristic extraction by first server 4, then carry out aspect ratio pair with the face in face sample database, second server 5 is for carrying out dilatation operation according to comparison result to sample database.Customer account management end comprises the first client 6 and the second client 7 etc., the operation such as is mainly used in manually increasing recognition of face sample database, deletes, changes, looks into.
Recognition of face has strict requirement to the facial image gathered in resolution, therefore need first to determine the adaptable minimum facial image resolution of algorithm according to recognition of face deviser, then sample worker estimates face imaging resolution according to camera interior and exterior parameter, finally selects suitable camera frame to set up an office and takes working point.Because the imaging resolution of Generic face recognizer requirement human face region is 96*96, in the present embodiment, adopt 1,000,000 grades of high-definition cameras.Camera can estimate according to the focal distance f of video camera (millimeter), monitoring width W (rice) and Sensor target surface size a (millimeter) from the distance U (rice) of guarded region central authorities: experimental formula is that U is approximately equal to f*W/a.General, video camera erection is highly overhead 2.5-4 rice, and monitoring width is about 3 meters, and the camera angle of pitch is within 15 degree, and the focal length of camera lens is 25-50mm.If when there is backlight, polarisation, camera is needed to support low-light (level) and the wide function such as dynamically.If for night-time scene, need to support auto iris.
Recognition of face sample collection method flow diagram is as shown in Figure 7:
First, determine the minimum face resolution restriction needed for recognition of face, such as require that face two eye pupil for recognition of face is not less than 50*50 pixel apart from distance, or require that the face of recognition of face is not less than 96*96 pixel in pixel.
Then, the multi-pose Face detection model of off-line training is loaded into face sample automated collection systems, such as 20*20 pixel; And the face characteristic extraction model in certain fixed resolution, such as 96*96 pixel.
Then, open face sample automated collection systems, and analyzing and processing is carried out to each frame in video.Processing procedure comprises the steps:
T1: video concentration, first utilizes the initial frame of video to set up the background model of each pixel or pixel region, and initial frame can choose the 1st frame or front 3 frames.After Background Modeling, calculating the foreground model of each pixel or pixel region, if find to differ greatly between the prospect of each pixel or pixel region and background model, is then that this pixel or pixel region give prospect mark.If the pixel containing prospect mark or pixel region possess certain scale, connected region such as containing prospect mark is more than 20*20 pixel, then export movable information mark, otherwise, then export without movable information mark, and upgrade without the pixel of movable information mark or the background model of pixel region.
T2: Face datection process, first concentrates the picture frame of module output to video or the areal map picture frame containing movable information carries out convergent-divergent and synthetic image pyramid; Then, do filtering operation thus the image that meets with a response by the multi-pose Face model of each layer pyramid diagram picture and off-line training, such as, the multi-pose Face model of off-line training is gang Haar wave filter, then allow each layer pyramid by this gang Haar wave filter; Finally, comprehensive analyze these filter responses, determine that whether the region of this motion identification is the concrete attitude of face and face.
T3: face tracking process, face tracking is mainly used in generating face movement locus.First on the face pixel region of face detection module output, face tracking model and face Track Initiation node is set up, face tracking model is made up of gang Haar wave filter etc., then, when the graphical analysis cycle arrives, around previous frame face pixel region, generate topography, allow this topography by face tracking model, then, according to the response of topography on model, the coordinate corresponding to peak response is joined the node tail of face track.
T4: face evaluation process, when the movement locus of certain face reaches image boundary, utilize the image information and face historical track information that cushion in video buffer pond, obtain the historigram picture frame of this face, and investigate the resolution of history face, attitude, sharpness and symmetry, provide the quality score of history facial image frame.This quality score is sorted, and exports those history facial image frames reached corresponding to recognition of face requirement of marking.
T5: face characteristic extraction process, first illumination pretreatment is carried out to history facial image frame and facial image frame is zoomed to face characteristic model restriction face resolution, such as 96*96 pixel; Then, set 40 Gabor cores, as 5 yardsticks, 8 directions, do convolution with these Gabor collecting images, then piecemeal obtains the feature of face; Then the Linear Redundancy degree between feature is reduced by Principle components analysis; Again by the discrimination between independent component analysis lifting feature; Finally, the face characteristic needed for recognition of face is obtained.
T6: face alignment with enter library module, each face characteristic in the face characteristic of history facial image frame and face Sample Storehouse is compared, as Euclidean distance etc., if the face characteristic of the feature of certain sample and collection " distance " is nearer, the face then gathered and feature thereof are appended in Sample Storehouse the information area depositing this face, otherwise, be then the face of collection and the newly-built information storage area of feature in face Sample Storehouse.
A kind of recognition of face sample collection method based on video mode that the present embodiment provides, complete in the automatic gatherer process of recognition of face sample, volunteer only need through position, control point, and sample collection worker only need ensure the normal and normal power supply of shooting point position of network, like this for sample collection worker, both saved time also laborsaving; Because the erection of camera fixed scene is according to face recognition algorithms design, and in gatherer process, camera interior and exterior parameter is constant, requires so the face of shooting meets recognition of face in resolution; Automatically face can be oriented from video, also the face of suitable recognition of face can be selected, automatic dilatation can also be carried out to original recognition of face Sample Storehouse according to the result of recognition of face, so, decrease the Data Enter of face Sample Storehouse managerial personnel to greatest extent.
Embodiment 3
The present embodiment provides a kind of recognition of face sample acquisition system based on video mode, as shown in Figure 8, comprising:
Image-region acquisition module, for obtain video picture frame in there is the image-region of movable information;
Face detection module, for detecting described image-region to obtain face pixel region;
Face tracking module, for following the tracks of containing face pixel region in the picture frame of face, obtains the historigram picture frame of same face;
Face characteristic acquisition module, for carrying out face characteristic extraction to the historigram picture frame of the high same face of quality, obtains the face characteristic needed for recognition of face;
Face characteristic memory module, for storing described face characteristic.
Described image-region acquisition module comprises:
Background Modeling submodule, sets up background model for utilizing the initial image frame in described video;
Foreground model sets up submodule, for setting up foreground model to current image frame;
Difference calculating sub module, for calculating the difference of described background model and described foreground model;
Image-region obtains submodule, is greater than the pixel region of predetermined threshold value, as the image-region with movable information for the difference obtained between foreground model and background model.
In described face detection module:
Image pyramid sets up submodule, for setting up image pyramid to current image frame;
Off-line model is loaded into submodule, for being loaded into each attitude faceform of off-line training;
Feature calculation submodule, for calculating the feature on each tomographic image pyramid;
Response calculating sub module, for calculating the response of described feature on each attitude faceform of off-line training;
Face pixel region obtains submodule, reaches the picture frame of Face datection threshold value for obtaining maximum response, extracts the attitude information of face pixel region and face in current image frame.
Described face tracking module comprises:
Submodule set up by face tracking model, for setting up face tracking model and face Track Initiation node;
Characteristic response calculating sub module, for calculating textural characteristics and the response of edge feature on face tracking model of face pixel region in current image frame;
Peak response process submodule, for adding the afterbody node of face track by the face pixel region coordinate in current image frame corresponding to peak response;
The historigram picture frame of same face obtains submodule, for when the face pixel region of current image frame arrives the border of picture frame, terminates to follow the tracks of, obtains the historigram picture frame from face Track Initiation node to the same face of face track afterbody node.
Described face characteristic acquisition module comprises:
Pre-service submodule, carries out illumination pretreatment for the picture frame high to quality;
Face characteristic calculating sub module, for calculating the face characteristic of face pixel region in the picture frame after illumination pretreatment;
Human face analysis submodule, for carrying out Principle components analysis and independent component analysis to described face characteristic;
Face characteristic obtains submodule, for obtaining the face characteristic needed for recognition of face.
Described face characteristic memory module comprises:
Face characteristic comparer module, for comparing each face characteristic in face characteristic described in face characteristic acquisition module and face Sample Storehouse;
Face characteristic sub module stored, if close for some face characteristics in described face characteristic and face Sample Storehouse, is then stored in this close face characteristic sample information Sample Storehouse by the face pixel region of described face characteristic and correspondence; If each face characteristic differs greatly in described face characteristic and face Sample Storehouse, then a newly-built face characteristic information Sample Storehouse stores the face pixel region of described face characteristic and correspondence.
A kind of recognition of face sample acquisition system based on video mode described in the present embodiment, first the image-region in the picture frame of video with movable information is obtained, then described image-region is detected to obtain face pixel region, then follow the tracks of containing face pixel region in the picture frame of face, obtain the historigram picture frame of same face, again face characteristic extraction is carried out to the historigram picture frame of the high same face of quality, obtain the face characteristic needed for recognition of face, finally store described face characteristic.The process of whole recognition of face sample collection does not need sample collection worker manual operation, and the speed of collection is fast, efficiency is high.
Those skilled in the art should understand, embodiments of the invention can be provided as method, system or computer program.Therefore, the present invention can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect.And the present invention can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disk memory, CD-ROM, optical memory etc.) of computer usable program code.
The present invention describes with reference to according to the process flow diagram of the method for the embodiment of the present invention, equipment (system) and computer program and/or block scheme.Should understand can by the combination of the flow process in each flow process in computer program instructions realization flow figure and/or block scheme and/or square frame and process flow diagram and/or block scheme and/or square frame.Can provide these computer program instructions to the processor of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing device to produce a machine, the instruction that the processor established by computing machine or other programmable data process is performed produces the device for realizing the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.
These computer program instructions also can be stored in can in the computer-readable memory that works in a specific way of vectoring computer or other programmable data processing device, the instruction making to be stored in this computer-readable memory produces the manufacture comprising command device, and this command device realizes the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.
These computer program instructions also can be loaded in computing machine or other programmable data processing device, make on computing machine or other programmable devices, to perform sequence of operations step to produce computer implemented process, thus the instruction performed on computing machine or other programmable devices is provided for the step realizing the function of specifying in process flow diagram flow process or multiple flow process and/or block scheme square frame or multiple square frame.
Although describe the preferred embodiments of the present invention, those skilled in the art once obtain the basic creative concept of cicada, then can make other change and amendment to these embodiments.So claims are intended to be interpreted as comprising preferred embodiment and falling into all changes and the amendment of the scope of the invention.
Claims (12)
1., based on a recognition of face sample collection method for video mode, it is characterized in that, comprise the steps:
Obtain the image-region in the picture frame of video with movable information;
Detect described image-region to obtain face pixel region;
Follow the tracks of containing face pixel region in the picture frame of face, obtain the historigram picture frame of same face;
Face characteristic extraction is carried out to the historigram picture frame of the high same face of quality, obtains the face characteristic needed for recognition of face;
Store described face characteristic.
2. the recognition of face sample collection method based on video mode according to claim 1, it is characterized in that, the image-region step in the picture frame of described acquisition video with movable information comprises:
The initial image frame in described video is utilized to set up background model;
Foreground model is set up to current image frame;
Calculate the difference of described background model and described foreground model;
The difference obtained between foreground model and background model is greater than the pixel region of predetermined threshold value, as the image-region with movable information.
3. the recognition of face sample collection method based on video mode according to claim 1 and 2, is characterized in that, the described image-region of described detection comprises to obtain face pixel region step:
Image pyramid is set up to current image frame;
Be loaded into each attitude faceform of off-line training;
Calculate the feature on each tomographic image pyramid;
Calculate the response of described feature on each attitude faceform of off-line training;
Obtain the picture frame that maximum response reaches Face datection threshold value, extract the attitude information of face pixel region and face in current image frame.
4. according to the arbitrary described recognition of face sample collection method based on video mode of claim 1-3, it is characterized in that, describedly carry out tracking step containing face pixel region in the picture frame of face comprise described:
Set up face tracking model and face Track Initiation node;
Calculate textural characteristics and the response of edge feature on face tracking model of face pixel region in current image frame;
Face pixel region coordinate in current image frame corresponding to peak response is added the afterbody node of face track;
If when the face pixel region of current image frame arrives the border of picture frame, terminate to follow the tracks of, obtain the historigram picture frame from face Track Initiation node to the same face of face track afterbody node.
5., according to the arbitrary described recognition of face sample collection method based on video mode of claim 1-4, it is characterized in that, the historigram picture frame of the described same face high to quality carries out face characteristic extraction step and comprises:
The picture frame high to quality carries out illumination pretreatment;
Calculate the face characteristic of face pixel region in the picture frame after illumination pretreatment;
Principle components analysis and independent component analysis are carried out to described face characteristic;
Obtain the face characteristic needed for recognition of face.
6., according to the arbitrary described recognition of face sample collection method based on video mode of claim 1-5, it is characterized in that, the described face characteristic step of described storage comprises:
Each face characteristic in described face characteristic and face Sample Storehouse is compared;
If some face characteristics are close in described face characteristic and face Sample Storehouse, then the face pixel region of described face characteristic and correspondence is stored in this close face characteristic sample information Sample Storehouse;
If each face characteristic differs greatly in described face characteristic and face Sample Storehouse, then a newly-built face characteristic information Sample Storehouse stores the face pixel region of described face characteristic and correspondence.
7., based on a recognition of face sample acquisition system for video mode, it is characterized in that, comprising:
Image-region acquisition module, for obtain video picture frame in there is the image-region of movable information;
Face detection module, for detecting described image-region to obtain face pixel region;
Face tracking module, for following the tracks of containing face pixel region in the picture frame of face, obtains the historigram picture frame of same face;
Face characteristic acquisition module, for carrying out face characteristic extraction to the historigram picture frame of the high same face of quality, obtains the face characteristic needed for recognition of face;
Face characteristic memory module, for storing described face characteristic.
8. the recognition of face sample acquisition system based on video mode according to claim 7, is characterized in that, described image-region acquisition module comprises:
Background Modeling submodule, sets up background model for utilizing the initial image frame in described video;
Foreground model sets up submodule, for setting up foreground model to current image frame;
Difference calculating sub module, for calculating the difference of described background model and described foreground model;
Image-region obtains submodule, is greater than the pixel region of predetermined threshold value, as the image-region with movable information for the difference obtained between foreground model and background model.
9. the recognition of face sample acquisition system based on video mode according to claim 7 or 8, is characterized in that, in described face detection module:
Image pyramid sets up submodule, for setting up image pyramid to current image frame;
Off-line model is loaded into submodule, for being loaded into each attitude faceform of off-line training;
Feature calculation submodule, for calculating the feature on each tomographic image pyramid;
Response calculating sub module, for calculating the response of described feature on each attitude faceform of off-line training;
Face pixel region obtains submodule, reaches the picture frame of Face datection threshold value for obtaining maximum response, extracts the attitude information of face pixel region and face in current image frame.
10., according to the arbitrary described recognition of face sample acquisition system based on video mode of claim 7-9, it is characterized in that, described face tracking module comprises:
Submodule set up by face tracking model, for setting up face tracking model and face Track Initiation node;
Characteristic response calculating sub module, for calculating textural characteristics and the response of edge feature on face tracking model of face pixel region in current image frame;
Peak response process submodule, for adding the afterbody node of face track by the face pixel region coordinate in current image frame corresponding to peak response;
The historigram picture frame of same face obtains submodule, for when the face pixel region of current image frame arrives the border of picture frame, terminates to follow the tracks of, obtains the historigram picture frame from face Track Initiation node to the same face of face track afterbody node.
11. according to the arbitrary described recognition of face sample acquisition system based on video mode of claim 7-10, and it is characterized in that, described face characteristic acquisition module comprises:
Pre-service submodule, carries out illumination pretreatment for the picture frame high to quality;
Face characteristic calculating sub module, for calculating the face characteristic of face pixel region in the picture frame after illumination pretreatment;
Human face analysis submodule, for carrying out Principle components analysis and independent component analysis to described face characteristic;
Face characteristic obtains submodule, for obtaining the face characteristic needed for recognition of face.
12. according to the arbitrary described recognition of face sample acquisition system based on video mode of claim 7-11, and it is characterized in that, described face characteristic memory module comprises:
Face characteristic comparer module, for comparing each face characteristic in face characteristic described in face characteristic acquisition module and face Sample Storehouse;
Face characteristic sub module stored, if close for some face characteristics in described face characteristic and face Sample Storehouse, is then stored in this close face characteristic sample information Sample Storehouse by the face pixel region of described face characteristic and correspondence; If each face characteristic differs greatly in described face characteristic and face Sample Storehouse, then a newly-built face characteristic information Sample Storehouse stores the face pixel region of described face characteristic and correspondence.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410720464.XA CN104463117B (en) | 2014-12-02 | 2014-12-02 | A kind of recognition of face sample collection method and system based on video mode |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410720464.XA CN104463117B (en) | 2014-12-02 | 2014-12-02 | A kind of recognition of face sample collection method and system based on video mode |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104463117A true CN104463117A (en) | 2015-03-25 |
CN104463117B CN104463117B (en) | 2018-07-03 |
Family
ID=52909133
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410720464.XA Active CN104463117B (en) | 2014-12-02 | 2014-12-02 | A kind of recognition of face sample collection method and system based on video mode |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104463117B (en) |
Cited By (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105426730A (en) * | 2015-12-28 | 2016-03-23 | 小米科技有限责任公司 | Login authentication processing method and device as well as terminal equipment |
CN105447459A (en) * | 2015-11-18 | 2016-03-30 | 上海海事大学 | Unmanned plane automation detection target and tracking method |
CN105654055A (en) * | 2015-12-29 | 2016-06-08 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Method for performing face recognition training by using video data |
CN105678289A (en) * | 2016-03-07 | 2016-06-15 | 谭圆圆 | Control method and device of unmanned aerial vehicle |
CN105930822A (en) * | 2016-05-11 | 2016-09-07 | 北京格灵深瞳信息技术有限公司 | Human face snapshot method and system |
CN107346417A (en) * | 2017-06-13 | 2017-11-14 | 浪潮金融信息技术有限公司 | Method for detecting human face and device |
CN107644219A (en) * | 2017-10-10 | 2018-01-30 | 广东欧珀移动通信有限公司 | Face registration method and related product |
CN108171207A (en) * | 2018-01-17 | 2018-06-15 | 百度在线网络技术(北京)有限公司 | Face identification method and device based on video sequence |
CN108256479A (en) * | 2018-01-17 | 2018-07-06 | 百度在线网络技术(北京)有限公司 | Face tracking method and device |
CN108491832A (en) * | 2018-05-21 | 2018-09-04 | 广西师范大学 | A kind of embedded human face identification follow-up mechanism and method |
CN109034247A (en) * | 2018-07-27 | 2018-12-18 | 北京以萨技术股份有限公司 | A kind of recognition of face sample extraction method of the higher degree based on track algorithm |
CN109145771A (en) * | 2018-08-01 | 2019-01-04 | 武汉普利商用机器有限公司 | A kind of face snap method and device |
WO2019033574A1 (en) * | 2017-08-17 | 2019-02-21 | 平安科技(深圳)有限公司 | Electronic device, dynamic video face recognition method and system, and storage medium |
CN109614841A (en) * | 2018-04-26 | 2019-04-12 | 杭州智诺科技股份有限公司 | Fast face detecting method in embedded system |
CN109859234A (en) * | 2017-11-29 | 2019-06-07 | 深圳Tcl新技术有限公司 | A kind of video human trace tracking method, device and storage medium |
WO2019134346A1 (en) * | 2018-01-08 | 2019-07-11 | 平安科技(深圳)有限公司 | Face recognition method, application server, and computer-readable storage medium |
CN110210290A (en) * | 2019-04-22 | 2019-09-06 | 平安科技(深圳)有限公司 | Face picture acquisition method, device and computer equipment |
CN110263704A (en) * | 2019-06-18 | 2019-09-20 | 浙江宇视科技有限公司 | Human face data acquisition method, device, server, video collector and medium |
CN110390229A (en) * | 2018-04-20 | 2019-10-29 | 杭州海康威视数字技术股份有限公司 | A kind of face picture screening technique, device, electronic equipment and storage medium |
CN110458026A (en) * | 2019-07-11 | 2019-11-15 | 成都市喜爱科技有限公司 | A kind of face acquisition method, device, system, equipment and medium |
CN110910549A (en) * | 2019-11-15 | 2020-03-24 | 江苏高泰软件技术有限公司 | Campus personnel safety management system based on deep learning and face recognition features |
CN111046806A (en) * | 2019-12-12 | 2020-04-21 | 天地伟业技术有限公司 | Heterogeneous image face recognition target library generation method |
CN111209818A (en) * | 2019-12-30 | 2020-05-29 | 新大陆数字技术股份有限公司 | Video individual identification method, system, equipment and readable storage medium |
CN111241868A (en) * | 2018-11-28 | 2020-06-05 | 杭州海康威视数字技术股份有限公司 | Face recognition system, method and device |
CN111652070A (en) * | 2020-05-07 | 2020-09-11 | 南京航空航天大学 | Face sequence collaborative recognition method based on surveillance video |
CN112164097A (en) * | 2020-10-20 | 2021-01-01 | 南京莱斯网信技术研究院有限公司 | Ship video detection sample acquisition method |
CN112487222A (en) * | 2020-11-30 | 2021-03-12 | 江苏正赫通信息科技有限公司 | Method for quickly searching and effectively storing similar human faces |
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 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101419670A (en) * | 2008-11-21 | 2009-04-29 | 复旦大学 | Video monitoring method and system based on advanced audio/video encoding standard |
CN102542492A (en) * | 2012-01-10 | 2012-07-04 | 清华大学 | System and method for evaluating effect of visual advertisement |
US8542941B2 (en) * | 2007-07-24 | 2013-09-24 | Nikon Corporation | Imaging device, image detecting method and focus adjusting method |
CN103605969A (en) * | 2013-11-28 | 2014-02-26 | Tcl集团股份有限公司 | Method and device for face inputting |
CN103942577A (en) * | 2014-04-29 | 2014-07-23 | 上海复控华龙微系统技术有限公司 | Identity identification method based on self-established sample library and composite characters in video monitoring |
-
2014
- 2014-12-02 CN CN201410720464.XA patent/CN104463117B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8542941B2 (en) * | 2007-07-24 | 2013-09-24 | Nikon Corporation | Imaging device, image detecting method and focus adjusting method |
CN101419670A (en) * | 2008-11-21 | 2009-04-29 | 复旦大学 | Video monitoring method and system based on advanced audio/video encoding standard |
CN102542492A (en) * | 2012-01-10 | 2012-07-04 | 清华大学 | System and method for evaluating effect of visual advertisement |
CN103605969A (en) * | 2013-11-28 | 2014-02-26 | Tcl集团股份有限公司 | Method and device for face inputting |
CN103942577A (en) * | 2014-04-29 | 2014-07-23 | 上海复控华龙微系统技术有限公司 | Identity identification method based on self-established sample library and composite characters in video monitoring |
Non-Patent Citations (1)
Title |
---|
桑海峰等,: ""视频监控下的人脸跟踪与识别系统"", 《计算机工程与应用》 * |
Cited By (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105447459A (en) * | 2015-11-18 | 2016-03-30 | 上海海事大学 | Unmanned plane automation detection target and tracking method |
CN105447459B (en) * | 2015-11-18 | 2019-03-22 | 上海海事大学 | A kind of unmanned plane detects target and tracking automatically |
CN105426730A (en) * | 2015-12-28 | 2016-03-23 | 小米科技有限责任公司 | Login authentication processing method and device as well as terminal equipment |
CN105654055A (en) * | 2015-12-29 | 2016-06-08 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Method for performing face recognition training by using video data |
CN105678289A (en) * | 2016-03-07 | 2016-06-15 | 谭圆圆 | Control method and device of unmanned aerial vehicle |
CN105930822A (en) * | 2016-05-11 | 2016-09-07 | 北京格灵深瞳信息技术有限公司 | Human face snapshot method and system |
CN107346417A (en) * | 2017-06-13 | 2017-11-14 | 浪潮金融信息技术有限公司 | Method for detecting human face and device |
CN107346417B (en) * | 2017-06-13 | 2020-09-01 | 浪潮金融信息技术有限公司 | Face detection method and device |
WO2019033574A1 (en) * | 2017-08-17 | 2019-02-21 | 平安科技(深圳)有限公司 | Electronic device, dynamic video face recognition method and system, and storage medium |
CN107644219A (en) * | 2017-10-10 | 2018-01-30 | 广东欧珀移动通信有限公司 | Face registration method and related product |
CN107644219B (en) * | 2017-10-10 | 2021-04-02 | Oppo广东移动通信有限公司 | Face registration method and related product |
CN109859234A (en) * | 2017-11-29 | 2019-06-07 | 深圳Tcl新技术有限公司 | A kind of video human trace tracking method, device and storage medium |
WO2019134346A1 (en) * | 2018-01-08 | 2019-07-11 | 平安科技(深圳)有限公司 | Face recognition method, application server, and computer-readable storage medium |
CN108256479B (en) * | 2018-01-17 | 2023-08-01 | 百度在线网络技术(北京)有限公司 | Face tracking method and device |
CN108256479A (en) * | 2018-01-17 | 2018-07-06 | 百度在线网络技术(北京)有限公司 | Face tracking method and device |
CN108171207A (en) * | 2018-01-17 | 2018-06-15 | 百度在线网络技术(北京)有限公司 | Face identification method and device based on video sequence |
CN110390229A (en) * | 2018-04-20 | 2019-10-29 | 杭州海康威视数字技术股份有限公司 | A kind of face picture screening technique, device, electronic equipment and storage medium |
CN109614841A (en) * | 2018-04-26 | 2019-04-12 | 杭州智诺科技股份有限公司 | Fast face detecting method in embedded system |
CN109614841B (en) * | 2018-04-26 | 2023-04-18 | 杭州智诺科技股份有限公司 | Rapid face detection method in embedded system |
CN108491832A (en) * | 2018-05-21 | 2018-09-04 | 广西师范大学 | A kind of embedded human face identification follow-up mechanism and method |
CN109034247A (en) * | 2018-07-27 | 2018-12-18 | 北京以萨技术股份有限公司 | A kind of recognition of face sample extraction method of the higher degree based on track algorithm |
CN109034247B (en) * | 2018-07-27 | 2021-04-23 | 北京以萨技术股份有限公司 | Tracking algorithm-based higher-purity face recognition sample extraction method |
CN109145771B (en) * | 2018-08-01 | 2020-11-20 | 武汉普利商用机器有限公司 | Face snapshot method and device |
CN109145771A (en) * | 2018-08-01 | 2019-01-04 | 武汉普利商用机器有限公司 | A kind of face snap method and device |
CN111241868A (en) * | 2018-11-28 | 2020-06-05 | 杭州海康威视数字技术股份有限公司 | Face recognition system, method and device |
CN111241868B (en) * | 2018-11-28 | 2024-03-08 | 杭州海康威视数字技术股份有限公司 | Face recognition system, method and device |
CN110210290A (en) * | 2019-04-22 | 2019-09-06 | 平安科技(深圳)有限公司 | Face picture acquisition method, device and computer equipment |
CN110263704A (en) * | 2019-06-18 | 2019-09-20 | 浙江宇视科技有限公司 | Human face data acquisition method, device, server, video collector and medium |
CN110263704B (en) * | 2019-06-18 | 2021-07-20 | 浙江宇视科技有限公司 | Face data acquisition method, device, server, video acquisition device and medium |
CN110458026A (en) * | 2019-07-11 | 2019-11-15 | 成都市喜爱科技有限公司 | A kind of face acquisition method, device, system, equipment and medium |
CN110910549A (en) * | 2019-11-15 | 2020-03-24 | 江苏高泰软件技术有限公司 | Campus personnel safety management system based on deep learning and face recognition features |
CN111046806A (en) * | 2019-12-12 | 2020-04-21 | 天地伟业技术有限公司 | Heterogeneous image face recognition target library generation method |
CN111209818A (en) * | 2019-12-30 | 2020-05-29 | 新大陆数字技术股份有限公司 | Video individual identification method, system, equipment and readable storage medium |
CN111652070A (en) * | 2020-05-07 | 2020-09-11 | 南京航空航天大学 | Face sequence collaborative recognition method based on surveillance video |
CN112164097A (en) * | 2020-10-20 | 2021-01-01 | 南京莱斯网信技术研究院有限公司 | Ship video detection sample acquisition method |
CN112164097B (en) * | 2020-10-20 | 2024-03-29 | 南京莱斯网信技术研究院有限公司 | Ship video detection sample collection method |
CN112487222A (en) * | 2020-11-30 | 2021-03-12 | 江苏正赫通信息科技有限公司 | Method for quickly searching and effectively storing similar human faces |
CN112487222B (en) * | 2020-11-30 | 2021-11-30 | 江苏正赫通信息科技有限公司 | Method for quickly searching and effectively storing similar human faces |
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 |
Also Published As
Publication number | Publication date |
---|---|
CN104463117B (en) | 2018-07-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104463117A (en) | Sample collection method and system used for face recognition and based on video | |
Gai et al. | A detection algorithm for cherry fruits based on the improved YOLO-v4 model | |
Zhang et al. | Pedestrian detection method based on Faster R-CNN | |
CN103824070B (en) | A kind of rapid pedestrian detection method based on computer vision | |
CN103207898B (en) | A kind of similar face method for quickly retrieving based on local sensitivity Hash | |
CN105373135B (en) | A kind of method and system of aircraft docking guidance and plane type recognition based on machine vision | |
CN110210276A (en) | A kind of motion track acquisition methods and its equipment, storage medium, terminal | |
CN107844785B (en) | A kind of method for detecting human face based on size estimation | |
CN108037770A (en) | Unmanned plane power transmission line polling system and method based on artificial intelligence | |
CN106127145B (en) | Pupil diameter and tracking | |
CN107798313A (en) | A kind of human posture recognition method, device, terminal and storage medium | |
CN104166841A (en) | Rapid detection identification method for specified pedestrian or vehicle in video monitoring network | |
CN104537689B (en) | Method for tracking target based on local contrast conspicuousness union feature | |
Ren et al. | A novel squeeze YOLO-based real-time people counting approach | |
CN102043953A (en) | Real-time-robust pedestrian detection method aiming at specific scene | |
CN105046197A (en) | Multi-template pedestrian detection method based on cluster | |
CN109035330A (en) | Cabinet approximating method, equipment and computer readable storage medium | |
Mao et al. | Development of power transmission line defects diagnosis system for UAV inspection based on binocular depth imaging technology | |
CN109685045A (en) | A kind of Moving Targets Based on Video Streams tracking and system | |
CN103593679A (en) | Visual human-hand tracking method based on online machine learning | |
CN104615986A (en) | Method for utilizing multiple detectors to conduct pedestrian detection on video images of scene change | |
CN104346802A (en) | Method and device for monitoring off-job behaviors of personnel | |
CN104281839A (en) | Body posture identification method and device | |
CN105022999A (en) | Man code company real-time acquisition system | |
CN108960076A (en) | Ear recognition and tracking based on convolutional neural networks |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |