Handling Technology Research Of Industrial Robots Based On Visual Information

Vision is a fast and efficient means for human to access external information,Similarly, the camera of machine vision system is equivalent to the robot’s eyes,through pictures taken processing and feature extraction, we can target objectpositioning and relevant data information. Thus, machine vision can improveindependent ability of robots, thereby further improving the flexibility of robots. Withthe rapid development of society, the industrial robots have been more and more intothe industrial production, and because of the heavy and repeated handling work in theindustrial production, handling robots are increasingly being applied. In order toaccelerate the efficiency and flexibility of the handling robot, in this paper, a machinevision system is integrated into robot handling systems, which has developed a robotichandling system based on visual information.This article first analyzes the machine vision system, choose the lens radial orderaberrations pinhole camera model after comparing the lens distortion impact, and usethe radial alignment constraint method to solve the internal and external parameters ofthe camera. According to the research projects requirements and experimentalconditions, image acquisition system based on VC++is developed. Finally, weightedaverage method is used to gray image after comparing several methods of the imagegray. After comparing trials, piecewise linear transformation and median filteringalgorithm is adopted for image pre-processing, and so it has a better edge retentive. Inthe characteristic parameter extraction, Otsu law of image segmentation is adopted,Then Canny operator is used to extract the edges of objects. Finally, algorithm isimproved based on the two strings intersect characteristics of round, and the center ofmass of the object is extracted successfully.Then the robot handling control software is studied. First, the principle ofcommunication between the computer and the robot is studied, including the analysisof communication frame formats in communication protocol, error checking methods, and sending or receiving the agreement of command file. After comparing the fourmethods of using VC++programming language developing serial communication,using third-party software to write communication class is selected. Second, add themachine vision system and a robotic handling system together, robot handling systembased on visual information is developed, and the focus is the development of handlingcontrol system. The features and applications of Motocom32programming software ofmotoman robot is analyzed, and handling control software is developed on the base ofthis software. Finally, connect visual system and handling control system together, so arobot handling system based on visual information is built, and the system interface isdesigned, which allows the system to be operated reasonably and conveniently.From the three main areas of machine vision systems, communication protocols,handling control software, a robotic handling system based on visual information isdesigned and developed, image acquisition, processing, extracting characteristicparameters and handling control integration system comes ture. After test debug, therobot can perform tasks in accordance with the procedure correctly.