摘要
在彩色图像中,不同的颜色所对应的灰度值可能相同,因此在灰度图像中检测不到的信息可能在彩色图像中被检测到。本文提出一种彩色积分图概念,在AdaBoost算法的分类器训练过程中分别得到彩色图像的人脸信息,并在分类器的后五层使用融合彩色信息的分类器分类。实验表明,本文方法增强了分类器的分类能力,提高了系统的正确检测率,降低了错误报警率。
In color images, different colors may correspond to the same gray value, so there are some information that may not be detected in gray image but in color image. This paper presents a new image representation called the ＂Color Integral Image＂. In the improve AdaBoost algorithm, the color information were obtained from the color images and five in the post - classifiers use the classifiers that integrated of color information to classify the color images. The experimental results show that the method improved the performance of classifiers, accurate detection rates and decrease the classifiers＇ false alarm rates.
出处
《微计算机应用》
2011年第3期7-11,共5页
Microcomputer Applications
