|国家预印本平台
首页|人脸检测算法中对肤色分割效果的加强

人脸检测算法中对肤色分割效果的加强

Reinforce to Segment the Skin Region in Face Detect

中文摘要英文摘要

本文介绍了基于肤色的人脸检测算法的原理与系统构成。实验采用了Anil K.Jain等人在YCbCr色彩空间上所建立的肤色模型,根据其在色度平面上的聚类结果,对待检图像进行肤色分割。此外提出在肤色分割前对图像光度进行高频提升滤波预处理步骤。该举有效摒除了与皮肤色泽相近的背景物体的干扰,加强了肤色分割效果,提高了最终的检测准确率。

In this paper, the theory of face detect algorithm based on skin color and a relevant system configuration are introduced. It’s experimented with skin color model in YCbCr color space by Anil K. Jain etal. , which segments skin and non-skin regions according to the cluster on hue plane. Another, A new method is addressed to pre-treat the luminance of images by high-frequency emphasis filter, which avoids the noise of background objects with the similar color as skin, and reinforce to segment facial regions. The new pretreatment increases the accuracy of face detect system.

樊宁

计算技术、计算机技术

肤色分割人脸检测算法高频提升滤波

skin segment face detect algorithm high-frequency emphasis filtering

樊宁.人脸检测算法中对肤色分割效果的加强[EB/OL].(2008-01-11)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/200801-384.点此复制

评论