CN113553965A - Person identity recognition method combining face recognition and human body recognition - Google Patents
Person identity recognition method combining face recognition and human body recognition Download PDFInfo
- Publication number
- CN113553965A CN113553965A CN202110852905.1A CN202110852905A CN113553965A CN 113553965 A CN113553965 A CN 113553965A CN 202110852905 A CN202110852905 A CN 202110852905A CN 113553965 A CN113553965 A CN 113553965A
- Authority
- CN
- China
- Prior art keywords
- human body
- face
- image
- recognition
- threshold value
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
- G06F18/256—Fusion techniques of classification results, e.g. of results related to same input data of results relating to different input data, e.g. multimodal recognition
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention relates to the technical field of video monitoring, in particular to a person identity identification method combining face identification and human body identification, which comprises the steps of firstly collecting clear face images and human body images of passengers at a security check place, then receiving video frames in monitoring equipment of a waiting room, identifying the face images in the video frames, simultaneously carrying out human body identification if the face identification is successful, judging that the passengers have clothes changing behaviors if human body characteristic comparison scores obtained by the human body identification are smaller than a clothes changing threshold value, and replacing the old human body images of the current passengers with the human body images corresponding to the successful current face identification; if the face recognition fails, then the human body recognition is carried out, and an alarm signal is sent only when the human body recognition fails, otherwise, the alarm signal is not sent. The invention combines the face recognition technology with the pedestrian recognition technology, and reduces the false alarm rate of character identity recognition under the non-matching scene. The invention has the advantages of reducing the false alarm rate and identifying the clothes changing behavior of passengers.
Description
Technical Field
The invention relates to the technical field of video monitoring, in particular to a person identity identification method combining face identification and human body identification.
Background
At present, the face recognition technology in the automatic safety inspection system of the high-speed rail is widely applied, and the face recognition technology is mainly applied to the aspect of comparing the face image and the identity card face image of the high-speed rail on site. With the continuous development of biological identification technology and the continuous expansion of the demand of high-speed rail stations for improving the efficiency of automatic security inspection systems, at present, more and more high-speed rail stations need to perform person authentication at a security inspection position in addition to a testimony comparison, and need to judge whether a person appearing in a video image is a corresponding passenger of a current train or not in other non-user-matched scenes such as a high-speed rail carriage, a waiting platform and the like.
The recognition rate of the face recognition technology in the prior art under the non-matching scene can be greatly influenced by various conditions such as illumination, face angles and the like, so that the possibility of higher false alarm can be caused, the high false alarm rate needs to be frequently verified by manual review, the user experience is reduced, and meanwhile, the labor cost of a station can be increased.
Disclosure of Invention
The invention aims to provide a person identity recognition method combining face recognition and human body recognition, which reduces the false alarm rate of person identity recognition under a non-fit scene by combining the face recognition technology with the pedestrian recognition technology and can recognize the clothes changing behavior of passengers.
The basic scheme provided by the invention is as follows: a person identity recognition method combining face recognition and human body recognition comprises the following steps:
s100: passenger information is obtained through a security check place, face detection and human body detection are carried out, and person ID labeling is respectively carried out on the detected face image and the detected human body image;
s200: receiving video frames transmitted back by other monitoring equipment, and extracting a face image and a human body image corresponding to the face image in the video frames;
s300: carrying out face recognition on the face image to obtain a face comparison score, judging the face comparison score and a preset face threshold value, and if the face recognition is successful, not sending an alarm signal; if the face recognition judgment fails, executing S400; the face threshold value comprises a first face threshold value, a second face threshold value and a third face threshold value, wherein the first face threshold value is larger than the second face threshold value, and the second face threshold value is larger than the third face threshold value;
s400: carrying out human body identification on the human body image to obtain a human body comparison score, judging the human body comparison score and a preset human body threshold value, if the human body identification judgment is successful, not sending an alarm signal, and if the human body identification judgment is failed, sending the alarm signal;
wherein, S300 includes:
s300-2: and if the face recognition is successful, carrying out human body recognition, comparing the human body image with the human body image corresponding to the face image acquired in the S100 to judge that the passenger has clothes changing behavior, deleting the human body image stored before, and storing the current human body image.
The principle and the advantages of the invention are as follows: according to the invention, human body recognition is added on the basis of obtaining accurate face information of a user, and when the face recognition is influenced by factors such as illumination, face angle and the like, the probability of false alarm caused by recognition error due to low face image quality can be remarkably reduced; meanwhile, after the face recognition is successful, the human body recognition result is judged, so that the clothes changing behavior of the passenger can be judged, the corresponding human body image library is updated in time, and the problem of clothes changing of the passenger can be effectively solved. Therefore, the invention has the advantages that: (1) the false alarm rate is reduced; (2) the clothes changing behavior of the passengers can be identified, and the problem of clothes changing of the passengers is effectively solved.
Further, the S100 includes:
s100-1: the face detection specifically comprises the steps of carrying out target detection data annotation on a recorded face image, framing the position of a face in the image, and then training a face detection model by utilizing annotation data.
Has the advantages that: the face information of the passenger can be recorded more accurately.
Further, the S100 includes:
s100-2: the human body detection specifically comprises the steps of carrying out target detection data labeling on a recorded human body image, framing the position of a human body in the image, and then training a human body detection model by using labeling data.
Has the advantages that: the method is favorable for recording the human body information of the passengers more accurately.
Further labeling the target detection data of the human face image, wherein the label comprises skin color, iris, mouth shape, face shape, nose shape and ear shape; the human body image target detection data labeling comprises the following steps: body type and dressing.
Has the advantages that: and the human face features and the human body features are subjected to multi-dimensional labeling, so that the feature extraction is more accurate.
Further, the video frame in S200 is a video frame within the last 30 frames in the video image transmitted back by the selected other monitoring device.
Has the advantages that: by selecting the video frames within the last 30 frames, the real-time performance of the video image can be ensured, and the storage space is saved.
Further, the S300-2 includes:
s300-2-1: storing the human body image corresponding to the human face image with the human face comparison score higher than a first human face threshold value in the video frame;
s300-2-2: and comparing the stored human body image with the current human body image of the passenger stored in the S100 to obtain a human body characteristic comparison score, comparing the obtained human body characteristic comparison score with a preset clothes changing threshold value, if the human body characteristic comparison score is smaller than the clothes changing threshold value, judging that the passenger has clothes changing behavior, replacing the current human body image with the human body image stored in the S100, otherwise, not processing.
Has the advantages that: the human body is identified through human body identification, and then the current image of the passenger can be updated at any time.
Further, the S400 further includes:
s400-1: if the face recognition fails, calling all human body images of passengers in the step S100, and carrying out the human body recognition with the human body images corresponding to the face images with unsuccessful face recognition, wherein the number of the human body images of each passenger in the step S100 is N;
s400-2: obtaining a human body identification result through human body identification in S400-1, if the number of human body images of which the human body comparison scores of the passengers are larger than the human body threshold value in the human body identification result exceeds the number of human body images of a passenger in S100 by more than N/2, judging that the alarm is false alarm, and storing the human body images of the passengers in the video frames; otherwise, the passenger is not considered as any passenger, and an alarm signal is sent out.
Has the advantages that: when the result of face recognition is poor, the false alarm rate is reduced through human body recognition, and the number of human body images of each passenger is N, which can be increased along with the increase of the database.
Further, the S100 further includes:
s100-3: the method comprises the steps of collecting ticket purchasing information of passengers and simultaneously carrying out association marking on the passengers.
Has the advantages that: by carrying out the association marking of the personnel in the same row, the retrieval of the association data of the personnel in the same row in the subsequent steps is facilitated.
Further, the S300 further includes:
s300-1: when the face image is subjected to face recognition, if the passenger has accompanying people, and the video frame shows that the face recognition of the passenger is successful, the face image of the passenger 'S accompanying people in S100-1 is taken and compared with the face image of the passenger' S accompanying people in the video frame.
Has the advantages that: by calling the information of the same-row personnel in the library and carrying out association matching, the identification efficiency can be improved, and the data occupation can be reduced.
Further, the S300 further includes:
s300-3: and if the face comparison score is smaller than the second face threshold value and the face comparison score is larger than the third face threshold value, identifying whether surrounding people are the same-person people of the passenger marked in the S100-3 in a correlated manner, if the identification is successful, not sending an alarm signal, and if the identification is failed, executing S400.
Has the advantages that: and reasonable threshold reduction is carried out on the result of face and human body recognition by using the information of the persons in the same row, so that the false alarm rate is further reduced.
Drawings
FIG. 1 is a block flow diagram of an embodiment of the present invention.
Detailed Description
The following is further detailed by way of specific embodiments:
the embodiment is basically as shown in the attached figure 1: a person identity recognition method combining face recognition and human body recognition comprises the following steps:
s100: passenger information is obtained through a security check place, face detection and human body detection are carried out, and person ID labeling is respectively carried out on the detected face image and the detected human body image;
s100-1: the face detection specifically comprises the steps of carrying out target detection data annotation on a recorded face image, framing the position of a face in the image, and then training a face detection model by utilizing annotation data;
s100-2: the human body detection specifically comprises the steps of carrying out target detection data labeling on a recorded human body image, framing the position of a human body in the image, and then training a human body detection model by using labeling data;
s100-3: the method comprises the steps of collecting ticket purchasing information of passengers and simultaneously carrying out association marking on the passengers.
In S100-1 and S100-2, target detection data annotation is realized by using a fast-RCNN algorithm, and the target detection data annotation of the face image comprises skin color, iris, mouth shape, face shape, nose shape and ear shape; the human body image target detection data labeling comprises the following steps: body type and dressing.
In the present embodiment, the person ID in S100 includes the name, identification card information, and the train number of the passenger.
S200: receiving video frames transmitted back by other monitoring equipment, and extracting a face image and a human body image corresponding to the face image in the video frames;
in S200, the video frame is a video frame within the last 30 frames of the video images transmitted back by other monitoring devices, and the video frame can be ensured to be the latest image by selecting the video frame of the last 30 frames, and the storage space can be saved.
S300: carrying out face recognition on the face image to obtain a face comparison score, judging the face comparison score and a preset face threshold value, and if the face recognition is successful, not sending an alarm signal; if the face recognition judgment fails, executing S400; in this embodiment, the Arcface algorithm is used for face recognition and the sphereid algorithm is used for human body recognition, wherein the Arcface algorithm and the sphereid algorithm are prior art and are not described herein in any further detail. Wherein the face threshold value includes a first face threshold value, a second face threshold value and a third face threshold value, specifically, the first face threshold value is 0.9, the second face threshold value is 0.8, the third face threshold value is 0.7, and the specific steps of S300 are:
s300-1: when the face image is subjected to face recognition, if the video frame shows that the passenger has accompanying people, the face recognition is carried out on the passenger to obtain a face comparison score, if the face comparison score is larger than a first face threshold value of 0.9, specifically 0.91 in the embodiment, the face recognition is judged to be successful, the face image of the accompanying people of the passenger in S100-1 is called and compared with the face image of the accompanying people of the passenger in the video frame, wherein after the step is successful in the recognition of the passenger, the accompanying people of the passenger are directly compared with the accompanying people of the passenger, the step that recognition software calls the face image of the whole database is reduced, the recognition efficiency is increased, and the face recognition can be carried out more quickly;
s300-2: and after the face recognition is successful, performing human body recognition, comparing the human body image with the human body image corresponding to the face image acquired in the S100 to judge that the user has clothes changing behavior, deleting the human body image stored before, and storing the current human body image. Wherein S300-2 comprises:
s300-2-1: storing the human body image corresponding to the human face image with the human face comparison score higher than the first human face threshold value in the video frame, wherein the passenger human face comparison score is 0.91 in the embodiment;
s300-2-2: and comparing the stored human body image with the current human body image of the passenger stored in the S100 to obtain a human body characteristic comparison score, comparing the obtained human body characteristic comparison score with a preset clothes changing threshold value, if the human body characteristic comparison score is smaller than the clothes changing threshold value, judging that the passenger has clothes changing behavior, replacing the current human body image with the human body image stored in the S100, otherwise, not processing. In this embodiment, the coat changing threshold is set to be 0.5, the stored human body images are compared with each other to obtain a human body feature comparison score, if the human body feature comparison score of the human body image is smaller than the set coat changing threshold of 0.5 at this time, in this embodiment, the human body feature comparison score is 0.4, it indicates that the user has a coat changing behavior, and the currently stored human body image is replaced with the previously stored human body image.
S300-3: if the face comparison score is smaller than the second face threshold value 0.8, it indicates that the recognition is failed, but at the same time, the face comparison score is larger than the third face threshold value 0.7, in this embodiment, the face comparison score is 0.75, it is recognized whether the surrounding people are the same-person people of the passenger associated with the mark in S100-3, if the recognition is successful, no alarm signal is sent, and if the recognition is failed, S400 is executed.
S400: performing human body recognition on the human body image to obtain a human body comparison score, and judging the human body comparison score and a preset human body threshold value, wherein in the embodiment, the preset human body threshold value is 0.8, if the human body recognition judgment is successful, no alarm signal is sent out, and if the human body recognition judgment is failed, an alarm signal is sent out, and the specific step of S400 is;
s400-1: if the face comparison score is smaller than the third face threshold, in this embodiment, the face comparison score is 0.6, calling all the human body images of the passengers in S100, and performing human body recognition on the human body images corresponding to the face images with unsuccessful face recognition currently, where N human body images of each passenger in S100 are taken;
s400-2: obtaining a human body comparison score through human body identification in S400-1, and comparing the human body comparison score with a preset human body threshold value, in the embodiment, comparing all human body images of the passenger in a video frame with the human body images of the passenger in S100, wherein the number of the human body images with the human body comparison score of the passenger being more than 0.8 exceeds the number of the human body images of the passenger in S100 by more than N/2, judging that the alarm is false alarm, and storing the human body images of the passenger in the video frame; otherwise, the passenger is not considered as any passenger, and an alarm signal is sent out.
The specific implementation process comprises the following steps:
the method comprises the steps that a passenger registers identity information, equipment collects face information and body information of the passenger, a face model and a body model are generated through algorithm fast-RCNN labeling, the equipment receives video frames transmitted back by other cameras, video images within the last 30 frames are intercepted, face recognition is conducted through an Arcface algorithm, if the face is not recognized, no response is made, after the face is detected successfully, body detection is conducted through a SphereRe algorithm, wherein whether a face comparison score is larger than a second face threshold value 0.8 or not and larger than 0.8 is judged, whether the face comparison score is larger than a first face threshold value 0.9 or not is continuously judged, if the face comparison score is larger than 0.9 or not is judged, body recognition is conducted, whether a body feature comparison score in a body image is smaller than a clothes changing threshold value or not is judged, wherein the clothes changing threshold value is 0.5, if the body feature comparison score in the body image is smaller than 0.5, all old body images of the person are deleted, and the newest body image of the person is stored, otherwise, the old human body image is not replaced.
If the face comparison score is between 0.8 and 0.9, the human body image comparison is not needed.
If the face comparison score is between 0.7 and 0.8, judging whether the person beside the person in the image is a related person accompanied with the person (the judgment method directly adopts face recognition), if so, not comparing the human body images, otherwise, comparing the human body images.
If the human face comparison score is less than 0.7, matching human body images, wherein the number of the human body images of each passenger is N, obtaining the human body image comparison score by comparing the human body images with the human body images of each passenger, if the number of the human body images with the human body comparison score being greater than the human body threshold value 0.8 is more than N/2, storing the human body images, judging that the passenger is the passenger of the train at the time through the identification, and the system cannot send an alarm signal, otherwise, judging that the passenger is not the passenger of the train at the time and sending the alarm signal if the number of the human body images with the human body comparison score being less than the human body threshold value 0.8 is less than N/2.
The foregoing are merely exemplary embodiments of the present invention, and no attempt is made to show structural details of the invention in more detail than is necessary for the fundamental understanding of the art, the description taken with the drawings making apparent to those skilled in the art how the several forms of the invention may be embodied in practice with the teachings of the invention. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several changes and modifications can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicability of the patent. The scope of the claims of the present application shall be determined by the contents of the claims, and the description of the embodiments and the like in the specification shall be used to explain the contents of the claims.
Claims (10)
1. A person identity recognition method combining face recognition and human body recognition is characterized in that: the method comprises the following steps:
s100: passenger information is obtained through a security check place, face detection and human body detection are carried out, and person ID labeling is respectively carried out on the detected face image and the detected human body image;
s200: receiving video frames transmitted back by other monitoring equipment, and extracting a face image and a human body image corresponding to the face image in the video frames;
s300: carrying out face recognition on the face image to obtain a face comparison score, judging the face comparison score and a preset face threshold value, and if the face recognition is successful, not sending an alarm signal; if the face recognition judgment fails, executing S400; the face threshold value comprises a first face threshold value, a second face threshold value and a third face threshold value, wherein the first face threshold value is larger than the second face threshold value, and the second face threshold value is larger than the third face threshold value;
s400: carrying out human body identification on the human body image to obtain a human body comparison score, judging the human body comparison score and a preset human body threshold value, if the human body identification judgment is successful, not sending an alarm signal, and if the human body identification judgment is failed, sending the alarm signal;
wherein, S300 includes:
s300-2: and if the face recognition is successful, carrying out human body recognition, comparing the human body image with the human body image corresponding to the face image acquired in the S100 to judge that the passenger has clothes changing behavior, deleting the human body image stored before, and storing the current human body image.
2. The method of claim 1, wherein the human identification method comprises the following steps: in S100, the method includes:
s100-1: the face detection specifically comprises the steps of carrying out target detection data annotation on a recorded face image, framing the position of a face in the image, and then training a face detection model by utilizing annotation data.
3. The method of claim 1, wherein the human identification method comprises the following steps: in S100, the method includes:
s100-2: the human body detection specifically comprises the steps of carrying out target detection data labeling on a recorded human body image, framing the position of a human body in the image, and then training a human body detection model by using labeling data.
4. The human identification method combining face recognition and human body recognition according to claim 2 or 3, wherein: the human face image target detection data labels comprise skin color, iris, mouth shape, face shape, nose shape and ear shape; the human body image target detection data labeling comprises the following steps: body type and dressing.
5. The human identification method combining face recognition and human body recognition according to claim 2 or 3, wherein: the video frame in S200 is a video frame within the last 30 frames in the video image transmitted back by the selected other monitoring device.
6. The method of claim 1, wherein the human identification method comprises the following steps: the S300-2 comprises:
s300-2-1: storing the human body image corresponding to the human face image with the human face comparison score higher than a first human face threshold value in the video frame;
s300-2-2: and comparing the stored human body image with the current human body image of the passenger stored in the S100 to obtain a human body characteristic comparison score, comparing the obtained human body characteristic comparison score with a preset clothes changing threshold value, if the human body characteristic comparison score is smaller than the clothes changing threshold value, judging that the passenger has clothes changing behavior, replacing the current human body image with the human body image stored in the S100, otherwise, not processing.
7. The method of claim 1, wherein the human identification method comprises the following steps: the S400 further includes:
s400-1: if the face recognition fails, calling all human body images of passengers in the step S100, and carrying out the human body recognition with the human body images corresponding to the face images with unsuccessful face recognition, wherein the number of the human body images of each passenger in the step S100 is N;
s400-2: obtaining a human body identification result through human body identification in S400-1, if the number of human body images of which the human body comparison scores of the passengers are larger than the human body threshold value in the human body identification result exceeds the number of human body images of a passenger in S100 by more than N/2, judging that the alarm is false alarm, and storing the human body images of the passengers in the video frames; otherwise, the passenger is not considered as any passenger, and an alarm signal is sent out.
8. The method of claim 1, wherein the human identification method comprises the following steps: the S100 further includes:
s100-3: the method comprises the steps of collecting ticket purchasing information of passengers and simultaneously carrying out association marking on the passengers.
9. The method of claim 8, wherein the human identification method comprises the following steps: the S300 further includes:
s300-1: when the face image is subjected to face recognition, if the passenger has accompanying people, and the video frame shows that the face recognition of the passenger is successful, the face image of the passenger 'S accompanying people in S100-1 is taken and compared with the face image of the passenger' S accompanying people in the video frame.
10. The method of claim 8, wherein the human identification method comprises the following steps: the S300 further includes:
s300-3: and if the face comparison score is smaller than the second face threshold value and the face comparison score is larger than the third face threshold value, identifying whether surrounding people are the same-person people of the passenger marked in the S100-3 in a correlated manner, if the identification is successful, not sending an alarm signal, and if the identification is failed, executing S400.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110852905.1A CN113553965B (en) | 2021-07-27 | 2021-07-27 | Person identity recognition method combining face recognition and human body recognition |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110852905.1A CN113553965B (en) | 2021-07-27 | 2021-07-27 | Person identity recognition method combining face recognition and human body recognition |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113553965A true CN113553965A (en) | 2021-10-26 |
CN113553965B CN113553965B (en) | 2023-09-12 |
Family
ID=78104609
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110852905.1A Active CN113553965B (en) | 2021-07-27 | 2021-07-27 | Person identity recognition method combining face recognition and human body recognition |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113553965B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110765955A (en) * | 2019-10-25 | 2020-02-07 | 北京威晟艾德尔科技有限公司 | Method for identifying human in video file |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105894625A (en) * | 2016-03-29 | 2016-08-24 | 深圳感官密码科技有限公司 | Intelligent barrier-free channel security identification system and method |
JP2017138680A (en) * | 2016-02-01 | 2017-08-10 | 富士通株式会社 | Image processing device and method |
CN109145742A (en) * | 2018-07-19 | 2019-01-04 | 银河水滴科技(北京)有限公司 | A kind of pedestrian recognition method and system |
CN110796079A (en) * | 2019-10-29 | 2020-02-14 | 深圳龙岗智能视听研究院 | Multi-camera visitor identification method and system based on face depth features and human body local depth features |
CN110956704A (en) * | 2019-11-21 | 2020-04-03 | 北京城建设计发展集团股份有限公司 | Novel efficient subway security inspection system and method |
CN111126158A (en) * | 2019-11-27 | 2020-05-08 | 中铁程科技有限责任公司 | Automatic ticket checking method, device and system based on face recognition |
US20200241874A1 (en) * | 2018-12-28 | 2020-07-30 | Shanghai Cambricon Information Technology Co., Ltd | Signal processing device, signal processing method and related products |
CN111967455A (en) * | 2020-10-23 | 2020-11-20 | 成都考拉悠然科技有限公司 | Method for comprehensively judging specified dressing based on computer vision |
CN112232113A (en) * | 2020-09-04 | 2021-01-15 | 北京旷视科技有限公司 | Person identification method, person identification device, storage medium, and electronic apparatus |
CN112749605A (en) * | 2020-02-26 | 2021-05-04 | 腾讯科技(深圳)有限公司 | Identity recognition method, system and equipment |
CN112784740A (en) * | 2021-01-21 | 2021-05-11 | 上海市公安局刑事侦查总队 | Gait data acquisition and labeling method and application |
CN112907773A (en) * | 2021-01-15 | 2021-06-04 | 佛山科学技术学院 | Intelligent attendance checking method and system based on motion detection and face recognition |
CN115862211A (en) * | 2022-11-22 | 2023-03-28 | 重庆电子工程职业学院 | Community security protection management and control system and method for trailing personnel |
-
2021
- 2021-07-27 CN CN202110852905.1A patent/CN113553965B/en active Active
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2017138680A (en) * | 2016-02-01 | 2017-08-10 | 富士通株式会社 | Image processing device and method |
CN105894625A (en) * | 2016-03-29 | 2016-08-24 | 深圳感官密码科技有限公司 | Intelligent barrier-free channel security identification system and method |
CN109145742A (en) * | 2018-07-19 | 2019-01-04 | 银河水滴科技(北京)有限公司 | A kind of pedestrian recognition method and system |
US20200241874A1 (en) * | 2018-12-28 | 2020-07-30 | Shanghai Cambricon Information Technology Co., Ltd | Signal processing device, signal processing method and related products |
CN110796079A (en) * | 2019-10-29 | 2020-02-14 | 深圳龙岗智能视听研究院 | Multi-camera visitor identification method and system based on face depth features and human body local depth features |
CN110956704A (en) * | 2019-11-21 | 2020-04-03 | 北京城建设计发展集团股份有限公司 | Novel efficient subway security inspection system and method |
CN111126158A (en) * | 2019-11-27 | 2020-05-08 | 中铁程科技有限责任公司 | Automatic ticket checking method, device and system based on face recognition |
CN112749605A (en) * | 2020-02-26 | 2021-05-04 | 腾讯科技(深圳)有限公司 | Identity recognition method, system and equipment |
CN112232113A (en) * | 2020-09-04 | 2021-01-15 | 北京旷视科技有限公司 | Person identification method, person identification device, storage medium, and electronic apparatus |
CN111967455A (en) * | 2020-10-23 | 2020-11-20 | 成都考拉悠然科技有限公司 | Method for comprehensively judging specified dressing based on computer vision |
CN112907773A (en) * | 2021-01-15 | 2021-06-04 | 佛山科学技术学院 | Intelligent attendance checking method and system based on motion detection and face recognition |
CN112784740A (en) * | 2021-01-21 | 2021-05-11 | 上海市公安局刑事侦查总队 | Gait data acquisition and labeling method and application |
CN115862211A (en) * | 2022-11-22 | 2023-03-28 | 重庆电子工程职业学院 | Community security protection management and control system and method for trailing personnel |
Non-Patent Citations (1)
Title |
---|
李胜广等: "人工智能在公安卡口中的应用", 《中国安防》, pages 80 - 85 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110765955A (en) * | 2019-10-25 | 2020-02-07 | 北京威晟艾德尔科技有限公司 | Method for identifying human in video file |
Also Published As
Publication number | Publication date |
---|---|
CN113553965B (en) | 2023-09-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP7040463B2 (en) | Analysis server, monitoring system, monitoring method and program | |
CN102945366B (en) | A kind of method and device of recognition of face | |
WO2015165365A1 (en) | Facial recognition method and system | |
CN109446936A (en) | A kind of personal identification method and device for monitoring scene | |
CN106845368A (en) | Airport boarding safety check based on recognition of face confirms system and method again | |
CN111680613B (en) | Method for detecting falling behavior of escalator passengers in real time | |
CN112634561A (en) | Safety alarm method and system based on image recognition | |
CN110738178A (en) | Garden construction safety detection method and device, computer equipment and storage medium | |
CN111460985A (en) | On-site worker track statistical method and system based on cross-camera human body matching | |
KR20190041704A (en) | A cctv divided management system and method it | |
CN106327876B (en) | A kind of fake-licensed car capture system and method based on automobile data recorder | |
CN114187561A (en) | Abnormal behavior identification method and device, terminal equipment and storage medium | |
CN113553965B (en) | Person identity recognition method combining face recognition and human body recognition | |
CN109800656B (en) | Positioning method and related product | |
CN113761964A (en) | Method for determining persons in the same group and related products | |
CN111055890B (en) | Intelligent detection method and detection system for railway vehicle anti-slip | |
CN112836557A (en) | Portrait identification method and system | |
CN110197722A (en) | AI-CPU system platform | |
CN111860100B (en) | Pedestrian number determining method and device, electronic equipment and readable storage medium | |
CN114283363A (en) | Passenger flow obtaining method and system based on face recognition and storage medium | |
CN113297976A (en) | Base station intrusion detection method and system based on deep learning | |
CN112149475B (en) | Luggage case verification method, device, system and storage medium | |
CN112149552A (en) | Intelligent monitoring method and device | |
CN112270218A (en) | Method and system for automatically identifying fingerprint | |
CN114627603B (en) | Warehouse safety early warning method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |