Research On The Punch Machine Protecting Technology Based On Computer Image Processing

In the modem industrial manufacturing, the developmental trend of working characteristic of the punch machine is high accuracy, super-speed and high automation. The quick rhythm and high-frequency of supplying materials of the punch machine makes operators suffer from tiredness both spiritual and physical, which leads to having hands and head wounded or disabled. As a result, it is necessary and urgent to design and install a safety equipment to protect operators in the punch machine. There are many protecting methods at home and abroad, such as using light electricity, infrared ray, mechanism, hydraulic pressure, etc., to protect this system. This paper brings up the research on the punch machine protecting technology based on computer image processing from a different point of view from other research methods. The paper provides a new algorithm based on theories of image processing, two color systems of RGB and HSI and texture analysis, workpieces and hands being regarded as recognizable objects. The results show that this method has the ability to distinguish between workpieces and hands, the results are acceptable. The contents of each chapter are described as follows:The first chapter expounds the significance of protecting technique of the punch machine and its situation and compares protecting techniques of the punch machine. Then the paper gives the possibility and advantages of this protecting technique, shows the research purpose, main contents and innovation points.The second chapter summarizes the system structure of image recognition, and also brings forward the practical model, hardware and software of this subject matter.The third chapter sums up the basic concepts of image recognition and characteristics pick-up, then analyzes the characteristics of the workpieces and hands respectively, and compare them.The forth chapter analyzes image pretreatment in three aspects: noise removing, cineration and restoration, and discusses deeply on the restoration algorithm of blurred images induced by relative motion.The fifth chapter generalizes three methods of image recognition, and sets forth the principles of fuzzy control, two color systems of RGB and HSI, and image texture. Applying these theories to recognition between workpieces and hands, the results of the image recognition are obtained. Finally, some other safety protecting methods are put forward.The sixth chapter discusses the main fruits and conclusions and presents some problems to be treated for further study.