CN113705422B - Method for obtaining character video clips through human faces - Google Patents
Method for obtaining character video clips through human faces Download PDFInfo
- Publication number
- CN113705422B CN113705422B CN202110977947.8A CN202110977947A CN113705422B CN 113705422 B CN113705422 B CN 113705422B CN 202110977947 A CN202110977947 A CN 202110977947A CN 113705422 B CN113705422 B CN 113705422B
- Authority
- CN
- China
- Prior art keywords
- frame
- face
- information
- video
- person
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 16
- 239000012634 fragment Substances 0.000 claims abstract description 33
- 238000001514 detection method Methods 0.000 claims description 9
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000012512 characterization method Methods 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
- Collating Specific Patterns (AREA)
Abstract
A method for obtaining a person video clip through a face generates a plurality of video clip information structures before inquiring a target person video clip, and the method for obtaining the target person video clip is quickened. The face appearing in the video and the video fragments appearing in the video are matched and stored, and each fragment is matched with the corresponding face characteristics. When the video clips of the target person are required to be acquired, the video clips of the target person can be obtained by comparing the face features of the target person face and the video person face clips to determine whether the person faces are the same person or not, and if so, the person faces are the same person. The video clips of all people can be marked in advance, and the face features of the target person and the video clips are only required to be compared later, so that the processing efficiency is greatly improved.
Description
Technical Field
The invention relates to the technical field of face recognition, in particular to a method for acquiring a character video clip through a face.
Background
Face recognition is a technique for identity matching by facial features of people. Under the application of various technologies, especially deep learning, it is possible to quickly detect a face from a picture or video stream and quickly match it with a face in a face library.
Face recognition is widely applied to various industries, such as entrance guard and theft prevention in security and protection, identity authentication in banks, even mobile phone face unlocking and the like. In the video related industry, if we need to obtain a video clip of a target person, obtaining the video clip through face comparison is a stable and low-false recognition mode. The traditional method is to compare the faces of the target person with the video frame by frame to see whether the target person exists or not, and obtain the video segment of the target person according to whether the target person exists or not.
Disclosure of Invention
In order to overcome the defects of the technology, the invention provides a method for acquiring the figure video clips through the human face, which marks the video clips of all the people in advance and improves the processing efficiency.
The technical scheme adopted for overcoming the technical problems is as follows:
a method for acquiring a character video clip through a human face comprises the following steps:
a) Taking frames from frame to frame or at intervals in a video stream, and obtaining coordinates of a face part in a frame picture through face detection;
b) Based on the coordinates of the face parts, capturing the images of the face parts to obtain face images, and preprocessing the face images;
c) Carrying out face recognition algorithm characterization on the preprocessed face picture to obtain a multidimensional feature vector;
d) Storing an information structure of a current frame, wherein the information structure of the current frame comprises a frame picture information structure, a frame fragment information structure and a video fragment structure;
e) Comparing the current frame picture information structure with the face information in the picture information structure of the previous frame, if the current frame and the previous frame have the same person, the current frame is the frame picture information of the same person, and if the current frame and the previous frame do not have the same person, the current frame is the frame picture information of a new person;
f) Updating frame picture information of the same person into a frame fragment information structure, updating a frame end position, adding frame picture information of a new person into the frame fragment information structure to enable the end position and the start position of the frame to be the same, and if the previous person information exists in the frame fragment information structure but the person information does not exist in the current frame, adding the person information into a video fragment structure to obtain a final stored video fragment structure;
g) And comparing the target person with the face information of the stored video clip structure, and if the comparison result is the same person, generating video clip information of the target person.
Further, the step of face detection in step a) includes:
a-1) detecting whether a face exists in the frame picture, if so, executing a step a-2), and if not, executing a step a-4);
a-2) carrying out face recognition on the frame pictures;
a-3) obtaining a face feature vector;
a-4) ending the face detection.
Further, the preprocessing in the step b) is cutting operation, deformation operation and normalization processing in sequence.
Further, the information structure of the current frame stored in step d) includes:
d-1) saving a frame picture information structure, wherein the frame picture information comprises face information, video names and frame position information;
d-2) saving a frame fragment information structure, wherein the frame fragment information comprises face information, video names, a frame starting position and a frame ending position;
d-3) storing a video clip structure, wherein the video clip comprises face information, a video name, a frame starting position and a frame ending position.
The beneficial effects of the invention are as follows: before inquiring the target person video clips, generating some video clip information structures to accelerate the method for obtaining the target person video clips. The face appearing in the video and the video fragments appearing in the video are matched and stored, and each fragment is matched with the corresponding face characteristics. When the video clips of the target person are required to be acquired, the video clips of the target person can be obtained by comparing the face features of the target person face and the video person face clips to determine whether the person faces are the same person or not, and if so, the person faces are the same person. The video clips of all people can be marked in advance, and the face features of the target person and the video clips are only required to be compared later, so that the processing efficiency is greatly improved.
Drawings
FIG. 1 is a face recognition flow chart of the present invention;
fig. 2 is a flow chart of video clip generation according to the present invention.
Detailed Description
The invention is further described with reference to fig. 1 and 2.
A method for acquiring a character video clip through a human face comprises the following steps:
a) And taking frames from frame to frame or at intervals in the video stream, and obtaining the coordinates of a face part in the frame picture through face detection.
b) And capturing the picture of the face part based on the coordinates of the face part to obtain a face picture, and preprocessing the face picture.
c) And carrying out face recognition algorithm characterization on the preprocessed face picture to obtain a multidimensional feature vector.
d) And storing the information structure of the current frame, wherein the information structure of the current frame comprises a frame picture information structure, a frame fragment information structure and a video fragment structure.
e) After the face recognition processing is completed, the face information of the current frame and the video information generate a picture information structure of the current frame. One face corresponds to one frame of picture information structure. The same frame picture may contain multiple faces, and thus multiple frame picture information structures. Comparing the current frame picture information structure with the face information in the picture information structure of the previous frame, if the current frame and the previous frame have the same person, the current frame is the frame picture information of the same person, and if the current frame and the previous frame do not have the same person, the current frame is the frame picture information of a new person.
f) And updating the frame picture information of the same person into a frame fragment information structure, updating the end position of the frame, adding the frame picture information of a new person into the frame fragment information structure to enable the end position and the start position of the frame to be the same, and adding the information of the person into a video fragment structure to obtain a final stored video fragment structure if the person information existing before in the frame fragment information structure but the person information does not exist in the current frame, which indicates that the person appears before but does not continue to appear.
g) And comparing the target person with the face information of the stored video clip structure, and if the comparison result is the same person, generating video clip information of the target person.
By the method, before the target person video clips are inquired, some video clip information structures can be generated, and the method for obtaining the target person video clips is quickened. The face appearing in the video and the video fragments appearing in the video are matched and stored, and each fragment is matched with the corresponding face characteristics. When the video clips of the target person are required to be acquired, the video clips of the target person can be obtained by comparing the face features of the target person face and the video person face clips to determine whether the person faces are the same person or not, and if so, the person faces are the same person. The video clips of all people can be marked in advance, and the face features of the target person and the video clips are only required to be compared later, so that the processing efficiency is greatly improved.
Example 1:
the face detection in step a) comprises the following steps:
a-1) detecting whether a face exists in the frame picture, if so, executing a step a-2), and if not, executing a step a-4);
a-2) carrying out face recognition on the frame pictures;
a-3) obtaining a face feature vector;
a-4) ending the face detection.
Example 2:
the pretreatment in the step b) is cutting operation, deformation operation and normalization treatment in sequence.
Example 3:
the information structure of the current frame stored in step d) includes:
d-1) saving a frame picture information structure, wherein the frame picture information comprises face information, video names and frame position information;
d-2) saving a frame fragment information structure, wherein the frame fragment information comprises face information, video names, a frame starting position and a frame ending position;
d-3) storing a video clip structure, wherein the video clip comprises face information, a video name, a frame starting position and a frame ending position.
Finally, it should be noted that: the foregoing description is only a preferred embodiment of the present invention, and the present invention is not limited thereto, but it is to be understood that modifications and equivalents of some of the technical features described in the foregoing embodiments may be made by those skilled in the art, although the present invention has been described in detail with reference to the foregoing embodiments. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (3)
1. A method for obtaining a video clip of a person through a face, comprising the steps of:
a) Taking frames from frame to frame or at intervals in a video stream, and obtaining coordinates of a face part in a frame picture through face detection;
b) Based on the coordinates of the face parts, capturing the images of the face parts to obtain face images, and preprocessing the face images;
c) Carrying out face recognition algorithm characterization on the preprocessed face picture to obtain a multidimensional feature vector;
d) Storing an information structure of a current frame, wherein the information structure of the current frame comprises a frame picture information structure, a frame fragment information structure and a video fragment structure;
e) Comparing the current frame picture information structure with the face information in the picture information structure of the previous frame, if the current frame and the previous frame have the same person, the current frame is the frame picture information of the same person, and if the current frame and the previous frame do not have the same person, the current frame is the frame picture information of a new person;
f) Updating frame picture information of the same person into a frame fragment information structure, updating a frame end position, adding frame picture information of a new person into the frame fragment information structure to enable the end position and the start position of the frame to be the same, and if the previous person information exists in the frame fragment information structure but the person information does not exist in the current frame, adding the person information into a video fragment structure to obtain a final stored video fragment structure;
g) Comparing the target person with the face information of the stored video clip structure, and if the comparison result is the same person, generating video clip information of the target person;
the information structure of the current frame stored in step d) includes:
d-1) saving a frame picture information structure, wherein the frame picture information comprises face information, video names and frame position information;
d-2) saving a frame fragment information structure, wherein the frame fragment information comprises face information, video names, a frame starting position and a frame ending position;
d-3) storing a video clip structure, wherein the video clip comprises face information, a video name, a frame starting position and a frame ending position.
2. The method for capturing video clips of a person via a face as in claim 1, wherein: the face detection in step a) comprises the following steps:
a-1) detecting whether a face exists in the frame picture, if so, executing a step a-2), and if not, executing a step a-4);
a-2) carrying out face recognition on the frame pictures;
a-3) obtaining a face feature vector;
a-4) ending the face detection.
3. The method for capturing video clips of a person via a face as in claim 1, wherein: the pretreatment in the step b) is cutting operation, deformation operation and normalization treatment in sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110977947.8A CN113705422B (en) | 2021-08-25 | 2021-08-25 | Method for obtaining character video clips through human faces |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110977947.8A CN113705422B (en) | 2021-08-25 | 2021-08-25 | Method for obtaining character video clips through human faces |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113705422A CN113705422A (en) | 2021-11-26 |
| CN113705422B true CN113705422B (en) | 2024-04-09 |
Family
ID=78654584
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110977947.8A Active CN113705422B (en) | 2021-08-25 | 2021-08-25 | Method for obtaining character video clips through human faces |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113705422B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102799871A (en) * | 2012-07-13 | 2012-11-28 | Tcl集团股份有限公司 | Method for tracking and recognizing face |
| CN103049459A (en) * | 2011-10-17 | 2013-04-17 | 天津市亚安科技股份有限公司 | Feature recognition based quick video retrieval method |
| CN103150546A (en) * | 2012-12-26 | 2013-06-12 | 冉阳 | Video face identification method and device |
| CN107333071A (en) * | 2017-06-30 | 2017-11-07 | 北京金山安全软件有限公司 | Video processing method and device, electronic equipment and storage medium |
| CN111126113A (en) * | 2018-11-01 | 2020-05-08 | 普天信息技术有限公司 | Method and device for processing face image |
| CN111460226A (en) * | 2020-04-01 | 2020-07-28 | 山东云缦智能科技有限公司 | Video character retrieval method and retrieval system based on deep learning |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7881505B2 (en) * | 2006-09-29 | 2011-02-01 | Pittsburgh Pattern Recognition, Inc. | Video retrieval system for human face content |
| CN107330408B (en) * | 2017-06-30 | 2021-04-20 | 北京乐蜜科技有限责任公司 | Video processing method and device, electronic equipment and storage medium |
-
2021
- 2021-08-25 CN CN202110977947.8A patent/CN113705422B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103049459A (en) * | 2011-10-17 | 2013-04-17 | 天津市亚安科技股份有限公司 | Feature recognition based quick video retrieval method |
| CN102799871A (en) * | 2012-07-13 | 2012-11-28 | Tcl集团股份有限公司 | Method for tracking and recognizing face |
| CN103150546A (en) * | 2012-12-26 | 2013-06-12 | 冉阳 | Video face identification method and device |
| CN107333071A (en) * | 2017-06-30 | 2017-11-07 | 北京金山安全软件有限公司 | Video processing method and device, electronic equipment and storage medium |
| CN111126113A (en) * | 2018-11-01 | 2020-05-08 | 普天信息技术有限公司 | Method and device for processing face image |
| CN111460226A (en) * | 2020-04-01 | 2020-07-28 | 山东云缦智能科技有限公司 | Video character retrieval method and retrieval system based on deep learning |
Non-Patent Citations (3)
| Title |
|---|
| Fast Search Algorithm for Short Video Clips from Large Video Database Using a Novel Histogram Feature;Feifei Lee等;《2008 International Conference on Computational Intelligence for Modelling Control & Automation》;20090724;1223-1227 * |
| Sioeye智能运动相机 开启人脸识别新时代;贺亮;《计算机与网络》;20180426;第44卷(第8期);29 * |
| 基于人脸识别的媒资视频检索技术的研究与实践;张志伟;《中国优秀硕士学位论文全文数据库 (信息科技辑)》;20131115(第11期);I138-734 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113705422A (en) | 2021-11-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102196686B1 (en) | Identity authentication method and device | |
| US7444007B2 (en) | Iris-based biometric identification | |
| US9613428B2 (en) | Fingerprint authentication using stitch and cut | |
| EP3364337A2 (en) | Persistent feature descriptors for video | |
| Zahedi et al. | License plate recognition system based on SIFT features | |
| JP2007317062A (en) | Person recognition apparatus and method | |
| BR9909623B1 (en) | process and apparatus for recognizing objects in an image frame as well as process for recognizing objects in a sequence of image frames. | |
| JP2012238121A (en) | Image recognition device, control method for the device, and program | |
| CN111339897B (en) | Living body identification method, living body identification device, computer device, and storage medium | |
| Winarno et al. | Anti-cheating presence system based on 3WPCA-dual vision face recognition | |
| CN105488486A (en) | Face recognition method and device for preventing photo attack | |
| US20080085039A1 (en) | Fingerprint matching method and apparatus using fingerprint minutiae and fingerprint binary images | |
| CN113705422B (en) | Method for obtaining character video clips through human faces | |
| US7436999B2 (en) | Data analysis device and data recognition device | |
| KR20080046490A (en) | Method for identifying face using montage and apparatus thereof | |
| CN111932755A (en) | Personnel passage verification method and device, computer equipment and storage medium | |
| Patterh | A biometric fusion based on face and fingerprint recognition using ANN | |
| CN111046804A (en) | Living body detection method, living body detection device, electronic equipment and readable storage medium | |
| EP3929806A3 (en) | Local encoding of intrinsic authentication data | |
| JP2005242432A (en) | Face authentication system and processing method for the system and program therefor | |
| CN111401167A (en) | Method for realizing biological recognition processing based on face recognition and palm vein recognition and corresponding system | |
| Bansal et al. | A fuzzfied approach to Iris recognition for mobile security | |
| KR20090031084A (en) | Face recognition system, the method of face recognition registration and confirmation | |
| KR20070097636A (en) | Method for detecting face robust to illumination change | |
| CN109145564A (en) | Control the method and device of mobile terminal |
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 | ||
| CB02 | Change of applicant information |
Country or region after: China Address after: 250000 West Zone, North 1st floor, S06 building, No. 1036, Langchao Road, high tech Zone, Jinan, Shandong Province Applicant after: Shandong Inspur Ultra HD Video Industry Co.,Ltd. Address before: 250000 West Zone, North 1st floor, S06 building, No. 1036, Langchao Road, high tech Zone, Jinan, Shandong Province Applicant before: Shandong Inspur Yunman Intelligent Technology Co.,Ltd. Country or region before: China Country or region after: China Address after: 250000 West Zone, North 1st floor, S06 building, No. 1036, Langchao Road, high tech Zone, Jinan, Shandong Province Applicant after: Shandong Inspur Yunman Intelligent Technology Co.,Ltd. Address before: 250101 west area, North 1st floor, S06 building, 1036 Chaochao Road, high tech Zone, Jinan City, Shandong Province Applicant before: SHANDONG YUNMAN INTELLIGENT TECHNOLOGY Co.,Ltd. Country or region before: China |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |