计算机视觉技术识别水果缺陷的一种新方法

利用计算机视觉技术识别水果表面缺陷，提出了分割缺陷和识别缺陷的新方法。该方法考虑了水果的形状特点，使不同灰度级的缺陷一次分割成功。在识别过程中利用水果的三维空间特征，实现了缺陷与果梗、花萼的区分。试验结果表明，该方法检测速度快，正确率高，能够满足计算机视觉进行在线检测的要求。

Study on Identifying Measurement About Default of Fruit in Computer Vision

The default on fruit face was identified by computer vision information. A new method to divide up and to identify default was presented. The character of fruit was considered. The division of default about different gray degree was successful at one time. The three dimensional character of fruit was utilized during identifying. The difference among default, stem end and calyces was found. By experiment, the speed of detection was very fast, and the accurate rate of detection is very high. The requirements of detection on line of computer vision.