基于改进Hough变换的堆叠工件定位方法

Positioning Method of Stacked Workpieces Based on Improved Hough Transform

摘要: 针对企业喷砂上料桁架机器人视觉系统中堆叠叉车工件难以识别问题，提出了一种改进的角约束的概率Hough变换直线检测算法，基于叉车工件边缘交叉点的判别依据制定直线分组规则，完成对堆叠叉车工件的识别与定位。结果表明，在堆叠不整齐、表面不一致等情况下，改进的概率Hough变换算法能准确检测边缘直线且精确定位不同型号叉车工件的中心点，为工业生产线实现对叉车工件的抓取提供支持，节约人力成本。

Abstract: In order to solve the problem that it is difficult to identify the stacked forklift workpieces in the visual system of the sandblasting and loading truss robot of enterprises, an improved angle con-strained probability Hough transform linear detection algorithm is proposed. Based on the judg-ment basis of the intersection of the forklift workpieces’ edges, a linear grouping rule is formulated to complete the identification and positioning of stacked forklift workpieces. The results show that the improved probability Hough transform algorithm can accurately detect the edge straight line and accurately locate the center point of different types of forklift workpieces under the conditions of irregular stacking and inconsistent surface, which provides support for industrial production lines to grasp forklift workpieces and saves labor costs.