Abstract: This paper presents an improved segmentation method for color pictures based upon color and texture, to segment the lion-eye area from lion-eye colorful pictures of pigs automatically. The research is based on RGB model. The selection of the initial point in region-growing arithmetic is improved. It is applied in color pictures. And a comparison of the improved algorithms is performed, which takes gray, color and both color and texture as eigenvectors. The results show that the improved region-growing arithmetic which grounds on R, G, B and texture has a perfect effect.

Key words: Image segmentation; Region-growing; Gray level concurrence matrix

摘要： 针对猪肉眼肌彩色图像中眼肌区域的提取，提出一种基于颜色和纹理特征的彩色图像分割方法，实现了眼肌区域的自动分割。该研究在RGB 彩色空间下进行，改进以往区域生长算法中种子点的选取方法，实现了基于彩色图像的区域生长算法，并比较以灰度为特征，以颜色为特征和以颜色与纹理相结合为特征进行区域生长的分割效果。研究结果表明，基于R、G、B 和纹理特征的改进区域生长算法具有较好的分割效果。

关键词: 图像分割；区域生长；灰度共生矩阵