一种主特征量自适应选择的多彩色空间图像分割方法研究

A Color Image Segmentation Method of Adaptive Selecting Primary Characters in Multiple Color Spaces

本文提出了一种基于多彩色空间自适应图像分割算法,在HSL和RGB空间通过平滑直方图至单峰,自适应地确定分割的主要特征量,再根据区域的邻接关系和色调距离进行区域合并,从而实现对图像的较好分割。

This paper proposes an adaptive algorithm of image segmentation using multiple colorspaces. This algorithm fixes on primary character by times of smoothing histogram to the only wave crest adaptively in HSL and RGB spaces, then makes a region merge based on the region spatial information and hue distance ,and gets a better result of the color image segmentation.