Design And Implementation Of A Transparent Workpiece Inspection System Based On CCD

At present, automatic monitoring and control on the molding surface is the key technology in automatic processing of transparent workpieces which need to be processed at high temperature or high pressure conditions in a closed environment.These works are completed by skilled workers relying on their experience of visual monitoring. The method of monitoring by experience with low efficiency and big error, cannot meet the requirements of modern manufacturing industry for the automation production.A monitoring system without contact and injuryto detect the molding process of transparent workpieceshas been investigated in this paper. When the size is close to a predetermined value of design, the detection system can real-time feedback the results to production control personnel, and provide the basis for the subsequent operation. This can be beneficial for the automatic control of the production process,also reduce the question of low efficiency and big errorresulted by monitoring byartificial experience.The basic principles of transparent workpiece forming and detection method have been analyzed in this paper, at the same time, the inspection system of transparent workpiece based on computer vision technology is proposed after the analyses of the detection system implementation from both sides of hardware and software. The overall design of the system is analyzed in detail, working principle and design principle of the system is also analyzed, as well as how to choose devices in the hardware system is introduced in the article. This paper studies the key technologies of laser spot image acquisition and processing.This paper introduces the application of OpenCV image processing technology such as the gray level, threshold segmentation, image filtering, erosion and dilation, contour detection and so on. The main research substance includes three aspects:firstly how to capture image data acquired by CCD camera through Ethernet interface and how to process and store correspondingly is studied. Secondly how to process laser facula edge by digital image processing technology OpenCV is investigated. Thirdly detection algorithm for spot is researched.In the process of forming a transparent workpiece, when detecting laser spot formed by shielding light path, the variation axis length of spot at each moment need to be calculated. The system is of good real-time and operatesspeedly,test results can meet the requirements. Compared the data of changing axial length and elevated height of surface moldingof a transparent workpiece, the axial length change thresholdcan be acquired to meet the conditions, thereby the goal of automatic detection can be achieved by setting threshold value for control system.