Research On Mechanical Parts Demension Detecting System

The visual inspection technology, which can extraordinarily reduce the inspection cost, greatly speed up of development and raise the efficiency of production, has become the topic of the day in the industrial online inspection in recent years which is widely used in the field of industrial inspection. In this paper, a size detecting system, which based on the machine vision technology and image process method has been developed. This system can be applied to the notched impact specimens'product line. The main research as follows:According to the request of the notched impact specimens'on-line detecting and the feature of machine vision system, on-line detecting system of the notched impact specimens'dimension and shape designed by using the technique of image acquisition, display, processing and analysis, which includes the structure design of the system, the configuration of the hardware, the comparison image of image processing, the design of the software and the final error analyzing. This paper have solved a series of key problems in the process of design which covers: the setting and calibration of the image acquisition system, the design of the image system lighting method, the research of the image processing method and image location. By the theoretical analysis and experimental test, the image processing procedure of the system has been designed. Firstly, the system needs carry out pre-process: the median filtering and average filtering of acquired image, next carrying out the threshold of filtered image, then performing morphology, such as open, close and so on. Next, The boundary of binary image is extracted. Dimension measurement of this notched impact specimen is implement after the detection of edge. Finally, the technique of match pattern should be used so that it can fix the location of the measure area on the measurement images. And then analyze the measurement error and provide a method of error correction. The process and measurement of images are all accomplished by using the powerful image processing software IMAQ Vision. Take LabVIEW as the platform finally, we have developed a human-computer interface for the notched impact specimen of system-run operation, management and display.