Face detection method and device adopting cascade classifier

Abstract

The invention discloses a face detection method and device adopting a cascade classifier, relating to the field of video monitoring. The face detection method comprises the following step of: carrying out face detection on rectangular windows with different positions and sizes in a current image to be detected by adopting a light type cascade-structure classifier to detect face windows, wherein each level of the light type cascade-structure classifier is a weak classifier. For a large quantity of windows to be detected in one image, relative to a scheme that a cascade-structure classifier is bound to eliminate partial non-faces after computing the results of all weak classifiers of a strong classifier, each weak classifier of the light type cascade-structure classifier can eliminate the partial non-faces, and therefore faster detection speed can be achieved, and the real-time detection of faces is further realized on megapixel high-definition videos.

Description

technical field [0001] The invention relates to the field of video monitoring, in particular to face detection technology in the field of video monitoring. Background technique [0002] With the improvement of security requirements, the commercial value of technologies such as people flow statistics, personnel feature recognition, and face recognition have begun to emerge, and they have gradually begun to be applied. Face detection, as the primary link of these tasks, has a very important role and significance , in recent years, researchers have invested a lot of time and energy in this field, devoting themselves to developing a fast and accurate face detection method. [0003] The method based on neural network is a very successful method in the earliest period of face detection, and its detection effect is very good, but the disadvantage is also very obvious: the detection speed is too slow. It was not until 2001 that Paul Viola et al. proposed the first real-time face de...

AI Technical Summary

Problems solved by technology

[0005] That is to say, for each window to be detected, the cascade structure classifier must calculate the results of all weak classifiers in a strong classifier before removing some non-faces, and an image often contains a large number of windows to be detected , so when the video resolution is higher, or even two million high-definition, the effect of real-time detection cannot be achieved

Further, in the case of extending the frontal face to the quasi-frontal face (that is, a face whose out-of-plane rotation does not exceed 60°), due to the more diverse training samples, the convergence speed in the iterative training process is slower during the training process using Haar-like features. , the detection speed of the final face detector is also slower, and it is more difficult to achieve quasi-frontal face detection

Images