Research On Visual Inspection Technology Of Ceramic Substrate Based On OpenCV

As a circuit insulation substrate,the finished size of ceramic substrate has a wide discrete band due to its material properties and processing technology.At present,ceramic substrate manufacturers also rely on traditional manual or semi-automated measurement methods in our country and ceramic substrates are classified according to dimensional tolerances.The detection speed of this method is slow,and manual measurement is easy to cause visual fatigue,leading to false detection or false detection,thus affecting the classification results.Therefore,using machine vision to detect ceramic substrates can avoid the error of manual measurement and reduce the production cost of enterprises.According to the requirements of ceramic substrate size detection,this paper developed a solution of ceramic substrate vision detection system based on OpenCV.The program is divided into hardware and software.Between them,the visual inspection system hardware is responsible for the acquisition of the ceramic substrate image,and automatically loading and sorting according to the detection result.The software part is responsible for the analysis and processing of the ceramic substrate image,the high-precision measurement of the size,the classification standard setting of the user interaction and the real-time display of detection results.This paper is divided into six chapters.In the first chapter,the source of the topic was introduced.The research status and development trend of visual inspection in industrial inspection at home and abroad were analyzed.The key technologies involved in visual inspection were summarized and the research ideas and main contents of this paper were introduced.In the second chapter,a set of solutions that ceramic substrate visual inspection system based on OpenCV were proposed according to the requirements of the project and the specific design of the visual inspection system development platform was explained.In the third chapter,the overall mechanical structure of the ceramic substrate visual inspection system was designed.The specific functions were introduced and the working platform,feeding mechanism and sorting mechanism were analyzed in detail.In the fourth chapter,the camera calibration and ceramic substrate image preprocessing algorithm was studied.The carama calibration method and calibration process were introduced.Besides the image calibration method,image enhancement,noise reduction algorithm and ceramic substrate edge extraction and refinement algorithm are studied to locate the edge of the ceramic substrate.In the fifth chapter,the size measuremnt method and experimental analysis of ceramic substrate based on sub-pixel subdivision wrer introduced.Through the research and analysis of the spatial moment sub-pixel subdivision algorithm,the precise positioning of the edge of the ceramic substrate was realized and the dimension measurement method was studied.Finally,the whole visiual detection system was tested several times.Through the analysis of the test results and the error,the measurement system has a maximum error of less than 2μm in the measurement range of 70mm,which meets the practical requirements.In the sixth chapter,the research contents of this topic were summarized and the research direction of further improvement was improved.