复杂背景下的人脸检测算法研究与实现

摘要

The face detection is a process that can find the faces in a picture,and determine the information such as location and size.Human face is a collection of information with extremely rich patterns,and it is the prime symbol for the human distinguishing from each other,understanding and memoring.The face detection in the field of human face image processing and computer vision occupies very important position.In view of color images based on complex background,this paper introduces several mainstream face detection algorithms.In the system design,a face detection algorithm based on skin color and feature fusion is proposed.This algorithm is a hierar⁃chical and coarse to precise detection method,and it can reduce the detection area step by step,so as to realize the detection and locali⁃zation of multiple faces.%人脸检测是指能够在一幅图像中找到其中存在的人脸，并能够确定其位置和大小等信息的过程。人脸是一个信息极丰富的模式集合，其是人类互相判别、认识、记忆的主要标志，人脸检测在人脸图像处理和计算机视觉领域占有极为重要的地位。针对复杂背景下的彩色图像，对目前主流的几种人脸检测算法进行了介绍，在系统设计中提出了一种基于肤色和特征融合的人脸检测算法。该算法是一种层次化、由粗到精的检测方法，通过一步步减小检测区域，从而实现多个人脸的检测与定位。