摘要
提出了一种基于彩色信息、直方图投影的人脸检测、跟踪算法。该方法适用于对时间要求较高、复杂背景下的运动图像人脸检测。测试结果表明,该方法在保护人脸大量的有用信息的情况下,大大改善了在复杂背景以及光照环境不佳情况下的人脸跟踪、识别效果,并具有计算量小的特点。
This paper proposes an algorithm for detecting and tracking human faces on the basis of chromatic histograms and backprojection. The algorithm is effectively adapted to realtime face detecting and tracking in case of high demand for time and complex background. The experimental results show that the algorithm not only can protect the useful information of human face and improve the success rate of detecting and tracking human face in a complex background or poor light condition,but it also has the merit of low computational cost.
出处
《贵州工业大学学报（自然科学版）》
CAS
2003年第3期46-49,53,共5页
Journal of Guizhou University of Technology(Natural Science Edition)
基金
国家自然科学基金(60272095)资助
