板料零件网格坐标图像特征提取算法研究

Research on Feature Extraction Algorithm of Grid Coordinates Images on Sheet Metal Parts

为进行板料零件冲压成形分析,采用网格坐标图像的视觉测量法进行变形区域塑性主应变分布的测量。该方法基于视觉测量变形区域应变分布;通过二维网格坐标图像的特征信息完成物体的三维重建,其中图像边界和特征点是常用的图像特征信息。在传统Freeman链码边界追踪算法的基础上进行改进可节省运算时间,跟踪后产生单像素宽度的轮廓边缘。在MFC中创建了一个基于OpenCV的应用程序,在单像素边界图像的基础上,利用OpenCV库函数实现对图像特征点的提取。试验结果表明,利用OpenCV实现特征点提取能省去不少编程的复杂工序,具有速度快和效率高的特点。

In order to carry out the forming analysis of sheet metal parts, the main strain distribution of plastic deformation region was measured using the visual image measurement method of grid coordinates. The key to measure the regional deformation and strain distribution based on visual measurement is to complete the reconstruction of three-dimensional objects by using the characteristics of two-dimensional grid coordinates images, and both image border and characteristic points are common image characteristic information. The traditional Freeman chain code of the border tracking algorithm was improved, which can save a great deal of operating time. After tracking, there are edges of single pixel width. A application program based on OpenCV was created in the MFC, image characteristic, points on single pixel width border were picked out using OpenCV. Test results show that using OpenCV to pick out feature points can omit the complex programming process, which has the characteristics of high speed and efficiency.