Research Of Defects Detection Technology Based On High Resolution PCB Image

Recently, PCB is developing with the orientation of high-density and multi-layers, and the machining quality of every layer would affect the quality of the board. To ensure the PCB's machining quality, every layer should be detected. The traditional detection method and algorithms can not meet the need of the modern PCB product which is high-density and multi-layers. However, on the contrary, the appearance of AOI (Automatic Optical Inspection) ameliorates all these disadvantages brought by the tradition methods significantly, what is more, AOI improves the precision and resolution, enhances the efficiency of manufacturing and decreases the costs of the production.Associated with the scientific develop trend and practical application, and based on a huge number of references as well as the analysis of traditional PCB image acquisition technology and processing system, the research work focused on defects detection methods and algorithms founded on the high resolution PCB image, which can be divided into several areas:(1) Through the thorough understanding of machine vision and the help of summary of defects detection technology, the AOI system was applied as the methodology to investigate the defects detection algorithm and realization based on high resolution PCB image.(2) Via the existed image scanning system platform, the research improved the design of acquisition and processing system, simultaneously, the brand-new algorithm and ideation, which was based on the physical features of PCB and lots of experience on experiment, attained the relatively high quality image signal.(3) When it came to the investigation of image pretreatment, the research work contracted with normal pretreatment advanced a high resolution PCB image pretreatment algorithm. And this evolutionary algorithm not only met the requirement of high calculation amount in high resolution image, but gained the satisfied outcomes.(4) Compared with the traditional algorithm in defects detection, the research penetrated into the geometrical characters of PCB, and then figured out a new defects detection algorithm concerned with geometrical features of the defects.(5) Developed the software on the platform VC++ 6.0. Combined the image acquisition system, the PCB defects detection system and the image processing system, and realized the application on the PCB defects detection.(6) Tested and amended the image processing algorithm according to the contrast of testing result and ideal result. Then discussed how to enhance performance and accuracy of the system after system's reliability and stability tested.