基于模糊C均值聚类的牛肉图像中脂肪和肌肉区域分割技术

Segmentation of fat and lean meat in beef images based on fuzzy C-means clustering

摘要: 介绍了一种自适应分割牛肉眼肌切面图像中脂肪和肌肉区域的图像处理技术。通过CCD摄像头获取以黑色平板为背景的牛肉眼肌切面彩色RGB图像。先根据彩色图像R分量的灰度直方图，利用最大方差自动取阀值法(OSTU)把黑色背景与整块牛肉图像分割开来；接着把处理后的图像变成灰度图像，用模糊C均值聚类算法(FCM)计算出牛肉脂肪像素和肌肉像素灰度值的聚类中心，以各个像素点灰度值与两个聚类中之间的绝对值距离来区分出图像中的脂肪和肌肉像素。结果表明，FCM方法是分割肌肉和脂肪区域的有效方法。

Abstract: A method of adapted segmentation of fat and lean meat in beef imaging was introduced. The color Red, Green, Blue(RGB) beef images using black pan as background was captured by CCD camera. Then, background was removed from the beef images by the Method of Maxium Classes Square Error(OSTU method) in the red color band of the RGB image. After beef images were converted from RGB to gray image, an adaptive segmentation method using a fuzzy C-means (FCM) clustering algorithm was developed to separate fat pixels from lean meat pixels.