Study on the Target Recognition and Location Technology of industrial Sorting Robot based on Machine Vision

Abstract

In order to improve the applications for an industrial sorting robot, it is necessary to increase its flexibility and control accuracy. The prerequisite is to automatically extract the multiple target positions accurately and robustly. The machine vision technology is an effective solution. Here an industrial robot arm is designed and set up for experiment simulation with machine vision. In order to reduce the influence of the size, deformation, and lighting etc., the target recognition and location method with fusion of scale invariant feature transform (SIFT) and moment invariants is developed. The experiments results showed that the developed image processing algorithms are robust, and the flexibility of the industrial robot can be improved by machine vision.