中文摘要：

针对梨表面缺陷的机器视觉检测问题，在对已有研究成果的分析和研究的基础上，论文采用形态学相加的方法实现梨图像的背景去除和表面缺陷提取；提出花萼、果梗与表面缺陷的区分方法；借助Matlab软件进行仿真算法的编程，通过作者设计开发的GraphicalUserInterface（GUI）界面，对三个品种的梨表面进行了缺陷检测仿真实验，成功提取了其中的表面缺陷信息，实验结果表明，作者提出的方法在多种梨的缺陷提取上通用性强、准确性高。

英文摘要：

In the method of machine vision-based pear surface quality detection, based on the analysis and investigation of the proposed research results, the method o{ simple morphology is adopted to detect defects and delete background, the distinguish method of calyx, stem and defect is presented. Matlab is adopted to program the simulation of the arithmetic, and by using the Graphical User Interface（GUI） de veloped by authors, three kinds of pear surface defects are detected successfully. The experimental results show that the method proposed by authors is universal and exact in process of detecting three kinds of pear.