• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2010, Vol. 32 ›› Issue (1): 64-66.doi: 10.3969/j.issn.1007130X.2010.

• 论文 • 上一篇    下一篇

基于颜色模型和纹理特征的彩色水果图像分割方法

  

  1. (1.南京农业大学工学院,江苏 南京 210031;2.南京农业大学信息科学技术学院,江苏 南京 210095)
  • 收稿日期:2008-07-13 修回日期:2008-11-10 出版日期:2010-01-18 发布日期:2010-01-18
  • 通讯作者: 210095 江苏省南京市南京农业大学信息科技学院计算机系 E-mail:xiezhonghong_cn@sina.com
  • 作者简介:谢忠红(1977-),女,江苏泰兴人,博士生,讲师,研究方向为图像处理与模式识别、计算机视觉和农业机器人。

A Method of Color Fruit Image Segmentation Based on the Color Model and the Texture Feature

  • Received:2008-07-13 Revised:2008-11-10 Online:2010-01-18 Published:2010-01-18

摘要:

本文介绍了RGB和HIS两种颜色模型,采用色调值H和色差值RG两个特征值,通过Otsu方法自动获取阈值对自然场景下彩色水果图像进行分割,并对实验结果进行了对比。然后,利用灰度共生矩阵分析了果实和背景的纹理特征,找到了区别果实和背景的纹理特征参数,对果实和背景颜色相似的图像进行分割。实验结果证明,纹理特征和颜色特征两者相辅相成,优势互补,对于包含复杂背景的青苹果图像分割效果比较好。

关键词: 纹理, 灰度共生矩阵, 分割

Abstract:

Firstly, the RGB model and the HIS model are introduced in this paper. The color fruit image under the  nature scene is segmented based on HUE and RG by using the threshold value gained according to Otsu's approach, and the results of segmentation are discussed. Secondly, the texture feature parameters to distinguish the fruit and the background are found by studying the texture differences between the fruit and the background using a gray level cooccurrence matrix. The color apple image with the similar background is segmented successfully. The experimental results show that the texture features and the color information help each other. The segmentation results are very good when the two methods are used in a green apple image including a complex background.

Key words: texture;gray level cooccurrence matrix;segmentation

中图分类号: