High-flux police dynamic face acquisition comparison method and system

Abstract

The invention discloses a high-flux police dynamic face acquisition comparison method and system. The method comprises the following steps: 1, rapidly detecting and snapshooting dynamic faces; 2, performing background system analysis; 3, performing three-dimensional rectification restoration; and 4, performing image preprocessing, feature extraction and comparison. According to the high-flux police dynamic face acquisition comparison method and system related to by the invention, for such features as many police portrait data image types, large quality differences, large data scale, diversified application scenes and the like, through a high-accuracy and rapid portrait feature extraction and comparison method, in a portrait comparison process, image deviations generated by the faces due to postures, decorations, illumination, focusing and the like are overcome, rapid and accurate comparison of face features is realized, a face identification technology is enabled to play an important role in actual application, and the method and system have application prospect.

Description

technical field [0001] The present invention relates to a dynamic face collection and comparison method and system, in particular to a high-throughput police dynamic face collection and comparison method and system, as well as its application in intelligent sensing channel doors. Background technique [0002] The video acquisition module in the perception channel of the prior art is mostly completed by the front-end network camera, and the collected video is encoded by H.264, and the video code stream is transmitted to the back-end analysis server, and the video code stream is decoded by the analysis server , and then perform intelligent video analysis on the decoded video image. The existing technology has obvious defects: 1. The video has gone through a process of encoding and decoding, and the video quality has a certain loss; 2. The process of video decoding needs to consume more computing resources of the back-end analysis server, and all video intelligent analysis Con...

AI Technical Summary

Problems solved by technology

The existing technology has obvious defects: 1. The video has gone through a process of encoding and decoding, and the video quality has a certain loss; 2. The process of video decoding needs to consume more computing resources of the back-end analysis server, and all video intelligent analysis Concentrating on the back-end server will also bring a large analysis delay, which is not suitable for applications with high real-time requirements

[0003] Most of the video acquisition modules in the perception channel of the prior art are based on the face detection algorithm. The defects are: 1. The face is easily camouflaged. If wearing a mask or sunglasses, the detection will fail, and the person’s face cannot be obtained. Data; 2. Face detection is easily affected by the angle of the face, and non-frontal faces are prone to detection failures resulting in missed reports and re-reports; 3. Face detection has high requirements for camera setup. The closer you are to the person, the better the effect, but It is relatively close and can easily cause pedestrians to be vigilant; 4. Since face detection only extracts face information, the obtained information only includes a single frontal face photo, and the amount of information is relatively small, which is not conducive to subsequent comprehensive applications

Existing technology cannot accurately and quickly extract and compare portrait features, making face recognition technology fail to play its due role