CN114359646A - Video analysis method, device, system, electronic equipment and medium - Google Patents
Video analysis method, device, system, electronic equipment and medium Download PDFInfo
- Publication number
- CN114359646A CN114359646A CN202011063257.3A CN202011063257A CN114359646A CN 114359646 A CN114359646 A CN 114359646A CN 202011063257 A CN202011063257 A CN 202011063257A CN 114359646 A CN114359646 A CN 114359646A
- Authority
- CN
- China
- Prior art keywords
- information
- video
- staff
- video data
- area
- 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
Landscapes
- General Factory Administration (AREA)
Abstract
The disclosure provides a video analysis method, a video analysis device, a video analysis system and electronic equipment. The method comprises the following steps: acquiring a plurality of video data acquired by at least two terminals in a first time range, wherein the terminal positions of the terminals are not completely the same; classifying the plurality of video data according to location information of each of the terminals; acquiring first information and second information of workers according to the plurality of video data, wherein the first information comprises worker identification information and dressing information; the second information comprises staff line-moving information and abnormal action information; and analyzing the behavior of the staff according to the first information and the second information.
Description
Technical Field
The present disclosure relates to the field of video analysis technologies, and in particular, to a method, an apparatus, a system, and an electronic device for video analysis.
Background
With the rapid development of computers and artificial intelligence, analysis of monitoring videos in various scenes can effectively help to process various problems. In the scene of a consumer, a monitoring video is mostly used for analyzing the consumer and monitoring an unexpected situation, and video data analysis is rarely used for providing evaluation data for the working quality of workers in the scene of the consumer, so that the working attitude and the working quality of the workers in the scene are evaluated.
Disclosure of Invention
The embodiment of the disclosure discloses a video analysis method, which comprises the following steps:
acquiring a plurality of video data acquired by at least two terminals in a first time range, wherein the terminal positions of the terminals are not completely the same;
classifying the video data according to the position information of each terminal, wherein the video data are respectively used for acquiring first information and second information of workers;
obtaining first information and second information of a worker according to the plurality of video data, wherein the first information comprises identification information and/or dressing information of the worker; the second information comprises staff line-moving information and/or abnormal action information;
and analyzing the behavior of the staff according to the first information and the second information.
For example, the plurality of video data are classified according to location information of each of the terminals, the location information including:
an access area and a work area.
For example, the first information and the second information of the staff are obtained according to the plurality of video data, and the first information comprises identification information and/or dressing information of the staff; the step that the second information comprises staff action information and/or abnormal action information comprises the following steps:
acquiring video data of the entrance and exit area, and acquiring first information of the staff according to the video data of the entrance and exit area;
and acquiring the video data of the working area, and acquiring second information of the staff according to the video data of the working area.
For example, the step of obtaining the video data of the entrance area and obtaining the first information of the staff according to the video data of the entrance area includes:
acquiring video data of the entrance and exit area to obtain a video image;
obtaining a face image in the video image according to the video image;
extracting face features in the face image, comparing the face features with a staff feature database, and identifying the staff in the video image;
and recording the time of identifying the staff by the video data, and obtaining the time of entering the workplace and the time of leaving the workplace of the staff to obtain the identification information of the staff.
For example, the step of obtaining the video data of the entrance area and obtaining the first information of the staff according to the video data of the entrance further includes:
acquiring video data of the entrance and exit area to obtain a video image;
obtaining a face image in the video image according to the video image;
extracting face features in the face image, comparing the face features with a staff feature database, and identifying the staff in the video image;
detecting a human body region in the video image, and obtaining a human body region corresponding to the human face region of the worker according to the worker feature database;
and extracting clothing features in the human body region of the worker, and matching the clothing features with a worker clothing feature database.
And obtaining the clothing feature matching result to obtain clothing matching information of the worker.
For example, the step of acquiring video data of a work area and acquiring second information of the staff according to the video data of the work area includes:
acquiring video data of a working area to obtain a plurality of frames of video images;
identifying and matching the face features in the multi-frame video images to obtain multi-frame pedestrian image coordinates;
and mapping the coordinates of the multi-frame pedestrian images to a real coordinate system to obtain the pedestrian action track.
For example, the step of mapping the coordinates of the multiple frames of pedestrian images to a real coordinate system to obtain the pedestrian action trajectory further includes:
when the pedestrian is judged to be a worker, the action track is marked by the name of the worker;
and when the pedestrian is judged to be a non-worker, the pedestrian track of the action track is identified by a digital ID.
For example, the action trajectory includes:
service area action tracks and non-service area action tracks.
For example, the second information includes staff service trajectory information and abnormal action information, and the abnormal action information includes:
expression specification detection and behavior specification detection.
For example, the behavior specification test includes the following steps:
the work area is divided into a contact area and a non-contact area.
For example, the staff attitude score is calculated and ranked according to the first information and the second information, and the steps further include:
acquiring the working time of a worker and the matching degree of decorations corresponding to the first information;
when the clothing matching degree is not smaller than a threshold value, judging that the clothing matching degree of the worker is 1;
acquiring the expression standard time length and the behavior standard times of the staff corresponding to the second information;
and when the expression standard duration is not less than a threshold value, judging that the expression standard degree of the staff is 1.
For example, video data collected by each terminal within a specific time range is acquired, and the specific time range is a specified working day.
The embodiment of the present disclosure further discloses a video analysis apparatus, including:
the video acquisition module is used for acquiring video data of each terminal;
the data generation module is used for carrying out content processing on each video to generate processing data;
and the data analysis module is used for analyzing the processing data to obtain the working attitude score and the ranking of the working personnel.
For example, the data generation module includes: the device comprises a video receiving unit, a video processing unit and a data sending unit.
The video receiving unit comprises an entrance area receiving subunit and a working area receiving subunit.
The video processing unit comprises an identification subunit, a dressing subunit, a line moving subunit and an abnormal action subunit.
The abnormal action subunit comprises an expression specification detection subunit and a behavior specification detection subunit.
Wherein the identification subunit and the clothing-wearing subunit are electrically connected with the doorway video receiving subunit; the action line subunit, the abnormal action subunit and the area video receiving subunit are electrically connected.
The video processing unit is electrically connected with the data sending unit.
The embodiment of the disclosure also discloses a video analysis system, which comprises a video acquisition terminal and a video analysis device, wherein,
the video acquisition terminal is configured to acquire video data;
the video analysis device is configured to process the video and output staff working attitude scores and rankings.
The embodiment of the present disclosure also discloses an electronic device, including: a processor, a memory, and a computer program stored on the memory and executable on the processor, wherein the processor implements one or more of the video analysis methods described above when executing the program.
Also disclosed are non-transitory computer-readable media comprising a computer program recorded thereon and capable of being executed by a processor, the computer program comprising program code instructions for implementing the method according to the above.
Drawings
Fig. 1 is a flow chart illustrating steps of a video analysis method provided by an embodiment of the present disclosure;
fig. 2 shows a distribution diagram of a video capture terminal provided by an embodiment of the present disclosure;
FIG. 3 illustrates a flow chart of clothing matching for a worker provided by an embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of a video analysis apparatus provided in an embodiment of the present disclosure;
fig. 5 shows a block diagram of an electronic device applying the video analysis method according to an embodiment of the present disclosure.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present disclosure more comprehensible, the present disclosure is described in further detail with reference to the accompanying drawings and the detailed description.
Referring to fig. 1, a flowchart illustrating steps of a video analysis method provided in an embodiment of the present disclosure is shown, which may specifically include the following steps:
step 101: a plurality of video data collected by at least two terminals in a first time range are obtained.
In the embodiment of the present disclosure, a plurality of video capture terminals are included in a consumer location, where the positions of the video capture terminals may be determined according to actual situations, but at least two sets of video capture terminals should be included, one set being installed in an entrance area and the other set being installed in a working area.
For example, referring to fig. 2, a distribution diagram of a video capture terminal provided by an embodiment of the present disclosure is shown.
Step 102: classifying the video data according to the installation position information of the video acquisition terminal, and respectively acquiring first information and second information of workers
In the embodiment of the disclosure, videos collected by the video collecting terminals in the entrance and exit areas are classified into entrance and exit videos, and the video data is used for acquiring identification information and dressing information of workers; the video collected by the working area video collecting terminal is classified into the working area video, and the video data is used for acquiring the line moving information and the abnormal action information of the staff.
It should be noted that the staff in this embodiment may be staff in various consumer scenarios, such as a salesperson in a mall, a serviceman in a bank, a serviceman in a catering consumption place, and the like. The application scene of the embodiment can be a consumption scene of a market, a bank business hall, a restaurant, an amusement park and the like.
As shown in fig. 2, the video data collected by the video collecting terminal in the door area is the video data of the entrance area and the exit area; 1. and 2, video data acquired by the video acquisition terminals at the positions 2 and 3 are the video data of the working area.
It is to be understood that the above examples are only examples set forth for a better understanding of the technical solutions of the embodiments of the present disclosure, and are not to be taken as the only limitations on the embodiments of the present disclosure.
Step 103: and obtaining the identification information and/or dressing information of the staff according to the video data.
For the identification information of the worker and the dressing information obtaining process, the following description of a specific implementation may be referred to.
And analyzing the entrance and exit videos of the consumption place, identifying the workers in the videos by using a face recognition technology, and obtaining identification information of the workers, wherein the identification information comprises the identity identification information and attendance time information of the workers.
Specifically, during a specified working day of a consumption place, videos of an entrance area and an exit area of the consumption place are processed, after the videos of the entrance area and the exit area of the consumption place are received, an image of the video is obtained, face detection is carried out on the image to obtain a face area in the image, face features are extracted, the extracted face features are compared with a face feature library of workers, and the workers in the video image are identified. After multi-frame video images are compared, selecting one frame of image with the earliest time of occurrence of the staff, and recording the video recording time corresponding to the image as t 1; selecting a frame of image with the latest time of occurrence of the same worker, and recording the video recording time corresponding to the image as t 2; and recording t1 as the work time of the worker, recording t2 as the work time of the worker, and calculating through t1 and t2 to obtain the work time of the worker. And recording the identified worker information as identity identification information, and recording the corresponding work attendance time, work leaving time and work duration of the worker as attendance identification information corresponding to the worker. The working time of the staff can also be obtained by analyzing the video in the working area. Analyzing videos in a working area to obtain the moving line track of a worker in the working area, obtaining a time period corresponding to at least one section of moving line track of the worker according to a starting time point and a disappearing time end point of the at least one section of moving line track of the worker, and calculating the time period of the at least one section of moving line track to obtain the in-store time length of the worker.
Step 104: and obtaining the line moving information and/or abnormal action information of the staff according to the video data.
The video data of the working area is analyzed, wherein the working area is divided into a service area and a non-service area, and after a worker enters the working area, the moving line track of the worker can be obtained through analysis of the video data obtained by the camera in the working area, and the time length information of the worker moving in different areas is obtained. If the activity or the stay time of the staff in the non-service area reaches a threshold value, the staff is considered to have a situation of work being lacked; and when the condition that the staff appears in the service area is detected, performing expression detection and assessment on the staff. The face of the worker in the video data of the working area is detected and analyzed, and the occurrence frequency data and the duration data of the face smile expression of the corresponding worker are obtained.
The work area still divide into contact district and non-contact district, when the staff has work to use communication tools such as cell-phone or other contact demands, the contact action should go on in the contact district, and the contact district can set up in the place that makes things convenient for managers supervision such as work scene proscenium, and the work area is as non-contact district except that other regions of contact district. Analyzing the video data of the working area, detecting the human body posture of the staff in the image, obtaining the times and the time length data of the staff using the mobile phone, combining the area position information of the staff when using the mobile phone, and recording the time length data of the staff in the contact area and the times and the time length data of the staff using the mobile phone in the non-contact area.
Preferably, real-time video captured by the camera is processed into a frame image, the image is subjected to a predictive tracking algorithm, all pedestrians in the frame image are identified, carrying out target tracking on the detected pedestrian of the previous frame in the next frame, carrying out distance matching on the detection frame of the current frame and the tracking frame of the previous frame, thereby determining the characteristic vector of the pedestrian in the current frame image, when the overlapping degree of the characteristic vector of the pedestrian in the current frame and the characteristic vector of the pedestrian in the previous frame is more than a preset threshold value, matching the pedestrians in the two frame images, determining the unique ID of the pedestrian in the current frame, when the pedestrian detected by the frame is not matched successfully, extracting the characteristics of the unmatched detection frame, and carrying out similarity calculation with the characteristics of the unmatched detection frame of the previous frame, when the similarity is greater than a certain threshold, the matching is considered to be successful, and the unique ID of the pedestrian is determined. And if the frame still has detection frames which are not successfully matched, the pedestrian corresponding to the detection frame is considered as a new pedestrian, and the pedestrian is endowed with a new ID. After the detection frame of the human body is detected, the coordinates (u, v) of each pedestrian in the image can be obtained through calculation, the physical positions of the boundaries of the video area are determined according to the physical positions of a plurality of characteristic points by acquiring the physical positions of the plurality of characteristic points of the boundaries of the video area through the visual positioning technology and the camera calibration method. For example, the region is a polygon region, and the plurality of feature points are a plurality of corner points of the region. The positions of the characteristic points can be obtained by a real-field measurement mode in the physical world, and can also be calculated according to the image positions and the position change relations of the characteristic points in the calibration images, so that the image pixels of the pedestrian are converted into two-dimensional coordinates projected on the ground in an actual coordinate system, and the two-dimensional coordinates of the pedestrian in a real world coordinate system are obtained, and the pedestrian track is formed.
When it is judged that the pedestrian is not a worker, identification is performed with a digital ID, which is unique. When the pedestrian is judged to be a worker, the face recognition information is associated with the track, the video recognizes the worker information by face recognition at the entrance and exit, meanwhile, the area coordinate of the face recognition in the recognition image is fixedly measured, the track with the minimum distance is selected to be matched with the face coordinate by performing distance calculation on the face coordinate (x1, y1) and the track coordinate (x2, y2), so that the information of the face is associated with the track, the track information of the corresponding worker is obtained, and the corresponding track information can be identified by the name of the worker.
It is understood that the detection of the pedestrian trajectory is a relatively mature technical solution in the field, and the embodiments of the present disclosure will not be described in detail herein.
After a worker enters a service area, after the sub-unit receives a video picture, firstly, a deep learning algorithm is used for detecting and positioning human faces and eyes, feature extraction is carried out on the detected human faces, the extracted features are input into a classifier, such as a deep learning classification frame of an SVM (support vector machine), a BOOST (BOOST transform) and the like, training is carried out in advance through a data set, and the human face features of the current worker are judged to be smiling faces or non-smiling faces in a classification mode.
The method based on human posture estimation is adopted to detect the behavior of the mobile phone, preferably, the existing deep learning algorithm model technology can be used for human body 2D key point detection and skeleton positioning, then a target detection technology such as FASTERCNN, YOLO, SSD and other algorithms is used near the hand joint point to detect that the target is the mobile phone, so as to determine whether the mobile phone exists near the hand; and constructing a space pose relation of a specific joint of the human body, monitoring the state of playing the mobile phone according to whether the space pose relation of the mobile phone and the human body is detected according to a target detection result, firstly determining the space pose of the upper limbs of the human body, determining that the upper limbs of the human body are within +/-100 degrees of a normal vector of the front face of the human body, secondly judging pose information of the whole face of the mobile phone and the human body so as to determine the relation between the mobile phone and the front view of the human body, and determining the state of using the mobile phone when the included angle between the mobile phone and the normal vector of the front face of the human body is within the range of 0-90 degrees. Wherein the normal vector of the front face of the human body is a normal vector corresponding to the front face of the human body as a plane within +/-100 degrees.
Step 105: and according to the video analysis information of the staff, performing score calculation and ranking on the working attitude of the staff.
The data received by the video analysis information comprises: identity information of workers and corresponding attendance time information; dressing information of the worker; the time node and the time length of the staff in the non-service area; the number of times and time of smiling face states of the worker; the number of times and the time that the worker uses the mobile phone in the non-contact area.
The score A corresponding to the identification information of the staff, the score B corresponding to the dressing information, the score C corresponding to the expression specification and the score D corresponding to the behavior specification are calculated according to the following formula:
Score＝α*A+β*B+γ*C+δ*D
the behavior specification score can be divided into a score E corresponding to the duration of the staff in the non-service area and a score corresponding to the number of times of using the mobile phone by the staff, and the corresponding formula is as follows:
D＝λ*E+θ*F
the scores corresponding to different information can be assigned with the same weight for calculation, and the weights of different working information can be adjusted according to different emphasis points in different evaluation modes, for example, in a bank scene, the identification information and the dressing information of the working personnel are emphasized, and the score weight matched with the identification information and the dressing information can be increased; in a catering scene, the duration and smile service of a worker in a non-service area are emphasized, and the weight occupied by the fraction corresponding to the information can be increased. The embodiments of the present disclosure do not limit this.
Optionally, the specific evaluation method may refer to the following calculation method:
if the attendance time of the staff is within the time range of going to work and the duration of the staff in the store meets the requirement, the attendance time meets the requirement and is represented by 0, otherwise, the attendance time does not meet the requirement and is represented by 1; when the dressing information of the workers is matched with the database, the dressing information is represented by 0, otherwise, the dressing information is represented by 1; calculating the real-time track position and time of a worker, calculating the time length of the worker staying in a non-service area, if the time length exceeds a threshold value, representing the time length by 1, or calculating the time length exceeding the threshold value and recording the time length; receiving smiling face states and time of workers in a service area, if the smiling face time is greater than threshold time or the number of smiling faces is greater than threshold number of times, regarding that the workers in smile assessment are in accordance with the regulations, and using 0 to represent the workers, otherwise, using 1 to represent or record smiling face duration or the number of smiling faces when the workers exceed the threshold time; and receiving the times of using the mobile phone by the staff in the non-contact area.
Optionally, the specific evaluation mode may refer to the following calculation method, and the evaluation results of each dimension are analyzed within a certain period of time, for example, during one day:
the total score for each worker is the sum of the scores for the above five dimensions. The work attitude ranking of the workers within one day can be obtained by ranking the total scores of the workers.
When the evaluation time length is a plurality of working days, for example, one month, the average score or the weighted average score of the working days in one month may be used for calculating and ranking the working attitude of the worker, the method of accumulating the working day scores may also be used for calculating and ranking, and similarly, the highest score or the lowest score of the working days may also be selected for calculating and ranking, which is not limited in the embodiment of the present disclosure.
The embodiment of the disclosure can analyze the working attitude of the staff by using the existing video, is beneficial to evaluating the working attitude of the staff in multiple dimensions, does not need to increase the workload of managers or the staff, and can help the managers to manage the staff, thereby improving the working efficiency and the work result.
Specifically, in the video analysis method provided in the embodiment of the present disclosure, the step 103 may include:
referring to fig. 3, a flowchart illustrating clothing matching of a worker provided in an embodiment of the present disclosure is shown, which may specifically include the following steps:
step 301: and receiving the entrance and exit videos.
Step 302: and carrying out face detection on the video image.
Step 303: comparing the face image obtained from the video image with the staff database
Step 304: and judging whether the face image in the video image is a worker.
Step 305: when the detected face image is a worker, the human body of the video image is detected to obtain the position of a human body detection area, and the position of the human body area of the worker is obtained by judging the face area in the position of the human body detection area. And detecting four elements of a hat, a hair accessory, clothes and shoes in the position to obtain the clothes data information of the staff.
Step 306: the worker clothing feature database stores templates corresponding to worker hats, worker hair ornaments, worker clothing and worker shoes, and obtained worker clothing data information is matched with template data in the standard database. The matching result is input to the video analysis unit.
Referring to fig. 4, a schematic structural diagram of a video analysis apparatus provided in the embodiment of the present disclosure is shown, which may specifically include the following modules:
and the video acquisition module 410 is configured to acquire video data acquired by the video acquisition terminal.
The data generating module 420 performs content processing on each of the videos to generate processed data.
The data generating module 420 includes a video receiving unit 421, a video processing unit 422, and a data transmitting unit 423.
Wherein the video receiving unit 421 includes:
the entrance and exit area receiving subunit 4211 is configured to obtain video data acquired by the entrance and exit area video acquisition terminals according to the position of the video acquisition terminal.
The working area receiving subunit 4212 is configured to obtain video data acquired by the working area video acquisition terminal according to the position of the video acquisition terminal.
The video processing unit 422 includes:
and the identification subunit 4221 is used for processing the video of the entrance and exit area to obtain the work information of the staff.
The dressing subunit 4222 is configured to process the video of the entrance area and the exit area to obtain clothing features of the worker, and compare the clothing features with the feature library to obtain dressing information of the worker.
And the line moving subunit 4223 is configured to process the video data of the work area to obtain line moving information of the worker.
The abnormal action subunit 4224 is configured to process the video data of the work area to obtain expression specification information and behavior specification information of the worker.
And the expression specification detection subunit 42241 is configured to process the video data of the work area, obtain facial expression features of the staff, and obtain the number of times and duration of expression specification by classification.
And the behavior specification detecting subunit 42242 is configured to process the video data of the work area, obtain human posture features of the staff, and obtain behavior non-specification times.
And the data analysis module 430 is used for analyzing the processing data to obtain the scores and the ranks of the workers.
Based on the same inventive concept, the embodiment of the disclosure also provides a video analysis system, which is composed of a video acquisition terminal and a video analysis device. The video acquisition terminal is configured to acquire video data, and different video terminals have different position information; the video analysis device is configured to process video data collected by the video collection terminal, perform analysis processing, and finally output scores and ranks of workers.
In another embodiment of the present disclosure, an electronic device is also provided. Fig. 5 shows a block diagram of an electronic device to which the above-described video analysis method may be applied, according to an embodiment of the present disclosure. As shown in fig. 5, the electronic device 500 may include: a processor 510, a memory 520 and a computer program stored on the memory and executable on the processor, which when executed by the processor implements the video analysis method in the above embodiments.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The video analysis method, the video analysis device, the video analysis system and the electronic device provided by the present disclosure are introduced in detail, and specific examples are applied in the text to explain the principle and the implementation of the present disclosure, and the description of the above embodiments is only used to help understanding the method and the core idea of the present disclosure; meanwhile, for a person skilled in the art, based on the idea of the present disclosure, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present disclosure should not be construed as a limitation to the present disclosure.
Claims (17)
1. A method of video analysis, comprising:
acquiring a plurality of video data acquired by at least two terminals in a first time range, wherein the terminal positions of the terminals are not completely the same;
classifying the video data according to the position information of each terminal, wherein the video data are respectively used for acquiring first information and second information of workers;
obtaining first information and second information of a worker according to the plurality of video data, wherein the first information comprises identification information and/or dressing information of the worker; the second information comprises staff line-moving information and/or abnormal action information;
and analyzing the behavior of the staff according to the first information and the second information.
2. The method of claim 1, wherein the classifying the plurality of video data according to location information of each of the terminals includes:
an access area and a work area.
3. The method of claim 2, wherein the obtaining of the first information and the second information of the staff from the plurality of video data comprises identification information and/or dressing information of the staff; the step that the second information comprises staff action information and/or abnormal action information comprises the following steps:
acquiring video data of the entrance and exit area, and acquiring first information of the staff according to the video data of the entrance and exit area;
and acquiring the video data of the working area, and acquiring second information of the staff according to the video data of the working area.
4. The method according to claim 3, wherein the step of obtaining the video data of the entrance area and obtaining the first information of the staff according to the video data of the entrance area comprises:
acquiring video data of the entrance and exit area to obtain a video image;
obtaining a face image in the video image according to the video image;
extracting face features in the face image, comparing the face features with a staff feature database, and identifying the staff in the video image;
and recording the time of identifying the staff by the video data, and obtaining the time of entering the workplace and the time of leaving the workplace of the staff to obtain the identification information of the staff.
5. The method of claim 3, wherein the step of obtaining video data of the entrance area and obtaining the first information of the staff according to the video data of the entrance further comprises:
acquiring video data of the entrance and exit area to obtain a video image;
obtaining a face image in the video image according to the video image;
extracting face features in the face image, comparing the face features with a staff feature database, and identifying the staff in the video image;
detecting a human body region in the video image, and obtaining a human body region corresponding to the human face region of the worker according to the worker feature database;
and extracting clothing features in the human body region of the worker, and matching the clothing features with a worker clothing feature database.
And obtaining the clothing feature matching result to obtain clothing matching information of the worker.
6. The method of claim 3, wherein the step of obtaining the video data of the work area and obtaining the second information of the staff according to the video data of the work area comprises:
acquiring video data of a working area to obtain a plurality of frames of video images;
identifying and matching the face features in the multi-frame video images to obtain multi-frame pedestrian image coordinates;
and mapping the coordinates of the multi-frame pedestrian images to a real coordinate system to obtain the pedestrian action track.
7. The method according to claim 6, wherein the step of mapping the coordinates of the multiple frames of pedestrian images to a real coordinate system to obtain the pedestrian action trajectory further comprises:
when the pedestrian is judged to be a worker, the action track is marked by the name of the worker;
and when the pedestrian is judged to be a non-worker, the pedestrian track of the action track is identified by a digital ID.
8. The method of claim 6, wherein the trajectory of action comprises:
service area action tracks and non-service area action tracks.
9. The method of claim 1, wherein,
the second information comprises staff service track information and abnormal action information, and the abnormal action information comprises:
expression specification detection and behavior specification detection.
10. The method of claim 9, wherein the behavior specification detection comprises the steps of:
the work area is divided into a contact area and a non-contact area.
11. The method of claim 1, wherein the first and second light sources are selected from the group consisting of,
analyzing the attitude of the staff according to the first information and the second information, wherein the steps further comprise:
acquiring the working time of a worker and the matching degree of decorations corresponding to the first information;
when the clothing matching degree is not smaller than a threshold value, judging that the clothing matching degree of the worker is 1;
acquiring the expression standard time length and the behavior standard times of the staff corresponding to the second information;
and when the expression standard duration is not less than a threshold value, judging that the expression standard degree of the staff is 1.
12. The method of claim 1, wherein
And acquiring video data collected by each terminal within a specific time range, wherein the specific time range is a specified working day.
13. A video analysis apparatus, wherein,
the device comprises a video acquisition module, a data generation module and a data analysis module;
the video acquisition module is used for acquiring video data of each terminal;
the data generation module is used for processing the content of each video to generate processing data;
and the data analysis module is used for analyzing the processing data to obtain the working attitude score and ranking of the working personnel.
14. The apparatus of claim 13, wherein,
the data generation module comprises: the device comprises a video receiving unit, a video processing unit and a data sending unit.
The video receiving unit comprises an entrance area receiving subunit and a working area receiving subunit.
The video processing unit comprises an identification subunit, a dressing subunit, a line moving subunit and an abnormal action subunit.
The abnormal action subunit comprises an expression specification detection subunit and a behavior specification detection subunit.
Wherein the identification subunit and the clothing-wearing subunit are electrically connected with the doorway video receiving subunit; the action line subunit, the abnormal action subunit and the area video receiving subunit are electrically connected.
The video processing unit is electrically connected with the data sending unit.
15. A video analytics system comprising: a video acquisition terminal and a video analysis device, wherein,
the video acquisition terminal is configured to acquire video data;
the video analysis device is configured to process the video and output staff working attitude scores and rankings.
16. An electronic device, comprising:
a processor, a memory, and a computer program stored on the memory and executable on the processor, the processor implementing the video analysis method of claim 1 when executing the program.
17. A computer readable storage medium having stored thereon computer instructions which, when executed by a processor, implement the video analytics method of claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011063257.3A CN114359646A (en) | 2020-09-30 | 2020-09-30 | Video analysis method, device, system, electronic equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011063257.3A CN114359646A (en) | 2020-09-30 | 2020-09-30 | Video analysis method, device, system, electronic equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114359646A true CN114359646A (en) | 2022-04-15 |
Family
ID=81090060
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011063257.3A Pending CN114359646A (en) | 2020-09-30 | 2020-09-30 | Video analysis method, device, system, electronic equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114359646A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114926054A (en) * | 2022-05-30 | 2022-08-19 | 安徽金源药业有限公司 | A5G intelligence factory data management system for health food processing |
-
2020
- 2020-09-30 CN CN202011063257.3A patent/CN114359646A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114926054A (en) * | 2022-05-30 | 2022-08-19 | 安徽金源药业有限公司 | A5G intelligence factory data management system for health food processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11354901B2 (en) | Activity recognition method and system | |
| CN107506684B (en) | Gait recognition method and device | |
| KR100886557B1 (en) | System and method for face recognition based on adaptive learning | |
| Salimi et al. | Visual-based trash detection and classification system for smart trash bin robot | |
| CN105740780B (en) | Method and device for detecting living human face | |
| JP6013241B2 (en) | Person recognition apparatus and method | |
| TWI505200B (en) | Method,system,computer program product and process for determining parts and associated attributes of an object in video | |
| CN109902628B (en) | Library seat management system based on vision thing networking | |
| Bedagkar-Gala et al. | Multiple person re-identification using part based spatio-temporal color appearance model | |
| CN109325964A (en) | A kind of face tracking methods, device and terminal | |
| CN109858371A (en) | The method and device of recognition of face | |
| CN106446754A (en) | Image identification method, metric learning method, image source identification method and devices | |
| KR101781358B1 (en) | Personal Identification System And Method By Face Recognition In Digital Image | |
| Kalas | Real time face detection and tracking using OpenCV | |
| CN110674680B (en) | Living body identification method, living body identification device and storage medium | |
| CN103150546A (en) | Video face identification method and device | |
| Sun et al. | Human recognition for following robots with a Kinect sensor | |
| Møgelmose et al. | Multimodal person re-identification using RGB-D sensors and a transient identification database | |
| Andiani et al. | Face recognition for work attendance using multitask convolutional neural network (MTCNN) and pre-trained facenet | |
| CN114581990A (en) | Intelligent running test method and device | |
| TW201828156A (en) | Image identification method, measurement learning method, and image source identification method and device capable of effectively dealing with the problem of asymmetric object image identification so as to possess better robustness and higher accuracy | |
| CN114359646A (en) | Video analysis method, device, system, electronic equipment and medium | |
| CN103020631A (en) | Human movement identification method based on star model | |
| CN109858308B (en) | Video retrieval device, video retrieval method, and storage medium | |
| JP2002208011A (en) | Image collation processing system and its method |
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 |