基于机器视觉的最大内接矩形快速检测算法

在进行皮革裁切、板材边角料再利用、PCB板布线等过程时都会遇到求解不规则图型最大内接矩形的问题;当前工厂的裁定手段普遍由人工估测来完成,这种作业方式不仅主观性强,而且效率低下;在研究了基于数字图像处理的遍历法、中心扩散法、遍历中心扩散法后,结合以上方法提出一种高效实用的最大内接矩形快速检测算法:边界排序生长法;实验结果证实:遍历法无法正确适用于凹多边形;中心扩散法虽然可适用于任意不规则图型,但其检测效果不够理想;遍历中心扩散法的检测结果较为理想,但该方法耗时巨大;而边界排序生长法在保证高效率的同时依旧可以拟合出理想的内接矩形.

A fast detection algorithm for maximum enclosed rectangle based on machine vision

The detection of irregular polygons with maximum enclosed rectangle is of great significance to the industrial scenes such as leather cutting, the reuse of board scraps and so on. At present, the method of adjudication is generally completed by manual estimation, this method has some problems such as strong subjectivity and low efficiency. After studying the traversing method, center diffusion method and traversal center diffusion method, a efficient and practical fast detection algorithm for maximum enclosed rectangle, boundary-sorting growth method, is proposed based on the above ideas. Experimental results show that the traversal method is not suitable for concave polygons. Although the central diffusion method can be applied to any graph, its detection effect is not ideal. The traversal center diffusion method has an ideal detection result, but it takes a lot of time. The boundary sorting growth method can ensure high efficiency and still obtain the ideal maximum enclosed rectangle.