CN102930261A - Face snapshot recognition method - Google Patents

Abstract

The invention discloses a face snapshot recognition method, which comprises the flowing steps: shooting a video; carrying out face tracking detection so as to obtain face images; carrying out assessment on each face image according to a face image quality assessment algorithm; screening out the optimal face image; retrieving a face photo library so as to obtain a set of face photo gallery similar to the optimal face image; and determining whether a face in the optimal face image exists in the face photo library. According to the invention, an effect of face recognition is achieved based on face snapshot, thereby improving the monitoring on various persons in monitoring areas.

Description

A kind of people's face is captured recognition methods

Technical field

The present invention relates to a kind of people's face and capture recognition methods.

Background technology

Video surveillance applications is in various safe and secret occasions, and in recent years, along with computing machine, network and image are processed, the develop rapidly of transmission technology, Video Supervision Technique also has significant progress.People's face is wherein captured technology, just refers to directly to capture people's face, as long as enter monitoring range, will grab record people face.But in the actual monitored situation, it is inadequate only just carrying out the candid photograph of people's face, also needs and can identify people's face.Therefore, how research capture people's face and identification is the problem that the applicant is devoted to solve.

Summary of the invention

The object of the present invention is to provide a kind of people's face to capture recognition methods, realize recognition of face on the basis that people's face is captured, thereby improve the control and monitoring to all kinds of personnel of guarded region.

The technical scheme that realizes above-mentioned purpose is:

A kind of people's face is captured recognition methods, comprises the following steps:

Step S1 is to entering personnel's capture video of surveyed area;

Step S2 carries out face tracking to video captured among the step S1 and detects, and obtains the facial image of each period;

Step S3 for each facial image that obtains among the step S2, estimates according to the quality of human face image evaluation algorithms;

Step S4, the quality of human face image evaluation result according to obtaining among the step S3 filters out the highest facial image of evaluation, as best facial image;

Step S5 retrieves in the human face photo storehouse for the best facial image among the step S4, obtains one group of human face photo collection similar to this best facial image;

Step S6 determines whether the people's face in the described best facial image is present in the middle of the described human face photo collection.

The invention has the beneficial effects as follows: the present invention is on the basis that people's face is captured, carry out successively quality of human face image evaluation, people's face picture retrieval and the contrast of people's face, thereby realization recognition of face, capture monitor procedure for the actual persons face orderly, a reasonable and effective face identification method is provided, improve the control and monitoring to all kinds of personnel of turnover guarded region.

Description of drawings

Fig. 1 is the process flow diagram that people's face of the present invention is captured recognition methods.

Embodiment

The invention will be further described below in conjunction with accompanying drawing.

See also Fig. 1, people's face of the present invention is captured recognition methods, comprises the following steps:

Step S1, to entering personnel's capture video of surveyed area, video can be divided into each period by the time;

Step S2 carries out face tracking to video captured among the step S1 and detects, and obtains the facial image of each period, namely captures out the facial image of each period for people's face;

Step S3 for each facial image that obtains among the step S2, estimates according to the quality of human face image evaluation algorithms;

Step S4 according to the quality of human face image evaluation result that obtains among the step S3, filters out the facial image of evaluation the highest (the most clear), as best facial image;

Step S5 retrieves in the human face photo storehouse for the best facial image among the step S4, obtains one group of human face photo collection similar to this best facial image, and namely each human face photo and the best facial image concentrated of this human face photo is the most approaching;

Step S6 by manual observation, determines whether the people's face in the described best facial image is present in the middle of the described human face photo collection, thereby realizes recognition of face.Can for different monitoring purposes, realize different targets by recognition of face; Such as, by recognition of face, confirm whether people's face of capturing is the staff that record is arranged, and if so, more can further confirm this staff's concrete identity.

So the present invention's energy Effective Raise is to all kinds of personnel's of turnover guarded region control and monitoring.

Above embodiment is only for explanation the present invention's, but not limitation of the present invention, person skilled in the relevant technique, in the situation that do not break away from the spirit and scope of the present invention, can also make various conversion or modification, therefore all technical schemes that are equal to also should belong to category of the present invention, should be limited by each claim.

Claims (1)

1. people's face is captured recognition methods, it is characterized in that, comprises the following steps:

Step S1 is to entering personnel's capture video of surveyed area;

Step S2 carries out face tracking to video captured among the step S1 and detects, and obtains the facial image of each period;

Step S3 for each facial image that obtains among the step S2, estimates according to the quality of human face image evaluation algorithms;

Step S4, the quality of human face image evaluation result according to obtaining among the step S3 filters out the highest facial image of evaluation, as best facial image;

Step S5 retrieves in the human face photo storehouse for the best facial image among the step S4, obtains one group of human face photo collection similar to this best facial image;

Step S6 determines whether the people's face in the described best facial image is present in the middle of the described human face photo collection.

Images