Gradient Direction Based Human Face Positioning Algorithm Applied in Complex Background

Face detection in complex background are vulnerable to light or other factors. This paper presents one face positioning algorithm based on the gradient direction. Firstly, smoothly preprocess the original image, then get the enhanced edge of grayscale image using sobel operator, binaries image, and finally set the position at the horizontal and vertical level with the characteristics of gradient direction of the binarized image. Experiments show that the algorithm for face recognition in complex background and positioning is fast and efficient. It also has the same effect on multiple face images of people.