Face detection method and device and computer readable storage medium

Abstract

The invention discloses a face detection method and device and a computer readable storage medium. The method comprises the steps of model training and face detection, wherein in the training process,multiple deep quadratic trees are constructed by the adoption of a positive sample image, a negative sample image and standard pixel difference features of the images, and then face detection tracking is performed through a random forest formed by the deep quadratic trees; in the training process, testing is performed every time the random forest is obtained to check whether a face can be correctly detected, and the image with a detection error is updated to a corresponding training set and subjected to continuous training learning till the random forest with a detection effect conforming toan expectation is obtained; and during detection, features of a to-be-detected image are input into the random forest, judgment is performed through each quadratic tree of the random forest, statistical analysis is performed on judgment results of all the quadratic trees, and whether the image is a face image is judged according to the statistical result. The method has the advantages that a detection model with high precision is obtained by use of a small quantity of training samples, and the method is particularly suitable for detection of a face with occlusion.

Description

technical field [0001] The invention relates to the fields of computer vision and image processing, in particular to a face detection method for detecting whether an image contains a face by using a face detector based on a random forest. Background technique [0002] With the rapid development of image processing technology and machine learning methods, face recognition-based applications are becoming more and more widespread, such as identity verification, intelligent transportation, security protection and other fields. Face detection is the first step in face recognition, and its accuracy is crucial. [0003] Face detection technology is a process of judging whether there is a face in the input image or video stream based on the characteristics of the face image. First, select an image feature, and perform feature extraction on the face sample image used for training; secondly, based on the extracted sample image feature, train a classifier to judge whether there is a f...

AI Technical Summary

Problems solved by technology

In some special occasions, there are certain obstacles in face detection, such as complex backgrounds, large differences in face shapes, or partial occlusion of faces caused by wearing accessories. It is difficult to design a universal face detection algorithm.

In addition, in order to achieve better detection results, a large number of face pictures are required to train the algorithm, and it is unrealistic to obtain a large-scale face picture set for face detection in a specific field

Images