【摘要】 针对金属切屑形态检测存在的问题,对C型屑、螺卷型屑和乱屑的形状特征进行了研究,提出了基于形状特征扩展的图像识别方法,即通过切屑图像的面积比特征粗分类法和扩展图像的欧拉数特征分类法,以MATLAB为平台,开发编制了相应的识别软件。经检验证明,本方法对3种典型的切屑形态识别具有可行性,其识别率达到98％以上。更多还原

【Abstract】 For the defects of supervision of chips’ shape, the figure - character of "C" chip, spiral chip and abnormal chip are studied, the new method of pattern recognition based on extend and transformed shape feature is brought forward, which is recognition method based on area contrast and recognition method by calculating Euler number of expanding image. The recognizing theoretical algorithm and computer program based on Matlab are worked out, and are proved to have a good feasibility and precision reaching to 98% by recognizing three types of chips’ shape.更多还原