Face detection method and device, computer device and computer readable storage medium

Abstract

The invention discloses a face detection method. The method comprises the following steps: constructing an image pyramid for an image to be detected; extracting aggregation channel characteristics ofeach layer of image of the image pyramid to obtain a characteristic pyramid of the to-be-detected image; obtaining a plurality of first detection frames of the to-be-detected image by using the firstsliding window, and classifying the first detection frames to obtain a plurality of candidate face frames; merging the candidate face frames; obtaining a plurality of second detection frames of the to-be-detected image by using a second sliding window, and classifying the second detection frames to obtain a plurality of candidate head-shoulder frames; combining the candidate header-shoulder framesand predicting a human face in the head-shoulder frames to obtain a predicted human face frame; and merging the merged candidate face frame and the predicted face frame to obtain a target face frame. The invention also provides a face detection device, a computer device and a readable storage medium. According to the invention, rapid high-detection-rate face detection can be realized.

Description

technical field [0001] The present invention relates to the technical field of computer vision, in particular to a face detection method and device, a computer device and a computer-readable storage medium. Background technique [0002] The commonly used methods for pedestrian capture are face detection, head-shoulder detection, and pedestrian detection. Due to the obvious and stable facial features, face detection has the highest detection rate and the lowest false detection rate among the three detection methods. However, the actual application scenario is more complex. Changes in face angle (head up, head down, side face), illumination changes (backlight, shadow), occlusion (sunglasses, masks, hats), etc. will reduce the detection rate of faces. Head-shoulder detection is to detect the head and shoulders as a whole. Because the head and shoulders are not as obvious and unique as the facial features, the detection effect is slightly worse than that of the human face. In ...

AI Technical Summary

Problems solved by technology

However, the actual application scenarios are more complex. Changes in face angles (head up, head down, side face), changes in illumination (backlight, shadows), occlusions (sunglasses, masks, hats), etc. will reduce the detection rate of faces.

Head-shoulder detection is to detect the head and shoulders as a whole. Because the head and shoulders are not as obvious and unique as the face features, the detection effect is slightly worse than that of the face.

In addition, head-shoulder detection generally uses edge features (HOG) or texture features (LBP), which are more complex and time-consuming to calculate.

Pedestrian detection generally requires the detection of the whole body. Pedestrians must all appear in the screen to be detected, which is often not met in actual scenes.

Images