Mechanical CAD parts recognition for industrial automation

Abstract

Object matching, counting and inspection are routine jobs done at several manufacturing plants, research laboratories, and other different companies. The detailed analysis of manufactured components is possible with information obtained from their inspection. For a large number of objects manual counting and inspection is a repetitive, difficult and time-taking process. The efficiency of overall object matching, counting and inspection process can be increased with industrial automation and it also minimizes resources and saves time. This paper presents a computationally efficient 3D computer vision based approach to recognize the Mechanical CAD parts. In this chapter features based industrial object detection techniques are implemented in MATLAB to recognize the presence of the industrial CAD parts in the query image.