Face tracking method and system based on deep learning
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A deep learning and face technology, applied in the field of image processing, can solve the problem of high robustness of face tracking, and achieve the effect of enhancing robustness, system robustness, and preventing face mistracking
Active Publication Date: 2020-04-07
HANGZHOU QUWEI SCI & TECH
View PDF7 Cites 6 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
The present invention is compatible with single-face and multiple-face tracking, and is not affected by the scene environment; the robustness of face tracking is high, and problems such as mistracking and lost tracking are solved; the mobile terminal of the present invention has strong real-time performance, the tracking speed can reach 100FPS, The method is very easy to implement on the mobile product side
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0037] like figure 1 As shown, this embodiment proposes a deep learning-based face tracking method, including:
[0038] S1. Obtain the start frame of the video stream as the current frame, and set n=0;
[0039] The video stream is composed of multiple frames of images, and the face tracking of the video stream is essentially the face detection and recognition of the multiple frames of images. In the face tracking method according to the specific embodiment of the present invention, the video stream is processed frame by frame. First, the start frame of the video stream is obtained through the image / video module, where n=0, and the video stream is obtained specifically through an image codec library, such as imageio library, OpenCV, and the like.
[0040] S2, determine whether the current frame n of the video stream satisfies n%N==0, if so, execute step S3, if not, execute step S4, wherein, N is the preset interval frame number;
[0041] In order to solve the problems of lar...
Embodiment 2
[0060] like figure 2 As shown, this embodiment proposes a face tracking system based on deep learning, including:
[0061] The image / video module is used to obtain the start frame of the video stream as the current frame, and set n=0;
[0062] The video stream is composed of multiple frames of images, and the face tracking of the video stream is essentially the face detection and recognition of the multiple frames of images. In the face tracking method according to the specific embodiment of the present invention, the video stream is processed frame by frame. First, the start frame of the video stream is obtained through the image / video module, where n=0, and the video stream is obtained specifically through an image codec library, such as imageio library, OpenCV, and the like.
[0063] The judgment module is used to judge whether the current frame n of the video stream satisfies n%N==0, if so, call the face detection module, if not, call the face verification module, where...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more
PUM
Login to view more
Abstract
The invention discloses a face tracking method and system based on deep learning, and the method comprises the steps: S1, obtaining a starting frame of a video stream as a current frame, and setting n= 0; s2, judging whether the current frame n of the video stream meets the condition that n% N is equal to 0 or not, if so, executing the step S3, and if not, executing the step S4, wherein N is a preset interval frame number; s3, performing face detection on the current frame, if a face is detected, outputting a face candidate box, and executing the step S4, otherwise, obtaining the next frame of the video stream as the current frame, setting n to be equal to 0, and executing the step S2; s4, performing face verification on the face candidate box, verifying whether the face candidate box contains a face or not, if so, outputting a face frame image, and executing the step S5, otherwise, obtaining a next frame of the video stream as a current frame, setting n to be equal to 0, and executing the step S2; s5, performing key point positioning on the face frame image, and calculating an external rectangular frame of a face key point; and S6, expanding the external rectangular frame to obtain an expanded rectangular frame, extracting a next frame of the video stream as a current frame, setting n = n + 1, setting the expanded rectangular frame as a face candidate frame, and executing thestep S2. The method is compatible with single-face and multi-face tracking, is not influenced by a scene environment, and is high in face tracking robustness and high in real-time performance.
Description
technical field [0001] The invention relates to the field of image processing, in particular to a deep learning-based face tracking method and system. Background technique [0002] In recent years, there have been more and more studies on face analysis. The so-called face analysis refers to the recognition of human expressions, locations, and identities based on the human face through computer vision and pattern recognition theory. Face tracking and face recognition are important links in face analysis, and through the cooperation of other links, the entire face analysis process can be effectively completed. Face analysis has come a long way, and it is well used in surveillance systems. Face tracking is the process of determining the motion trajectories and scale changes of one or more faces in a video stream or image sequence. Face tracking has great significance in the fields of image recognition analysis and target tracking, and has been the focus of attention in the fi...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.