基于多目视觉的变形钢板三维测量方法.pdfVIP

计算机测量与控制.201 1. 19( 的 |自动化测试! Computer Measurement Control • 2119 • 文章编号:1611-4598(2011)09 -2119 -03 中国分类号:TP391 文献标识码:A 基于多目视觉的变形钢板三维测量方法 史合龙， 61 毒系，嵌在宦， J ~童雯，剑1.华 (江苏科技大学，计算机科学与工程学院，江苏镇江 212003) 摘要:提出了一种使用 4 个相机测量变形钢板气维形状的方法;在没有纹理的钢板上画一些点，使用 4 个相机同时拍摄钢板3 利用 边缘检测算法找出钢板的边缘，利用特征点提取算法提取出 4 个图像中钢饭上的特征点，并找到 4 个图像上精确的特征点匹配关系，然 后测量出这些特征点的二维点z 最后，以这些气维点为种子点向外区域增长，从而得到钢板的三维形状s 实验对变形钢板进行气维测量， 得到了钢板的精确的士气雏形状，结果验证了该方法的有效性和可行性。 关键词:视觉5 三维测量，变形钢板 3D Measuremeot Method of Deformable Steel Surface Based 00 Multi - Visioo Shi Jinlong. Bai Suqin , Tan Youxia , Ding Jianwen. Liu Qinghua (School of Computer Science and Engineering. Jiangsu University of Science and Technology. Zhenjiang 212003. China) Abstract: This paper presents a method for measurement of deformable steel plate using four cameras. lf the plate is texture less. we need draw some points on the steel plate. First. four cameras are used to capture the images of steel plate ,Second. edge detection algorithm is adopted to find the edge of steel plate in images ,Third. feature points on images are extracted by feature extraction algorithm and the cor- respondences of feature points among the fours images are found. Acc咽rding to the matching of feature points. 3D points are measured. Fi- nally. region growing is implemented based on the measured 3D points. and 3D shape of steel plate is obtained. The performance and effec- tiveness are evaluated by simulated and real data experiments. Key words: computer vision; 3D measurement ,deformable steel plate 上画或投影一些随机纹理. 。 引言