Microscope Images Based Study On Inspection Algorithm Of PCB Surface Fault

Printed Circuit Board(PCB)is the carrier of all kinds of electron components and it tends to be higher density,higher integration,lighter,thinner and more wearable with development of electronic equipment.Therefore,it is important to study automatic defect detection in bare PCB.By taking AOI as basis,this thesis proposed a micrographic-image based detection algorithm in bare PCB with high speed and accuracy to detect defects in bare PCB.Firstly,this algorithm improves light source module,which makes il umina t io n intensity more average.Secondly,an image preprocessing regimen was proposed to reduce noise and stress edge features of the image through integrated application of Gaussian filtering,median filtering and Wavelet Filtering.Thirdly,an algorithm to realize a quick image mosaic of micrographic image was proposed and optimized for its real-time performance.Fourthly,the defect detection algorithm based on morphology was proposed to precisely recognize distinguishing bare defects in PCB through classification of defects in bare PCB.Therefore,Hence,the proposed algorit hm realizes image mosaic from micrographic and partial image to complete image in PCB with high density.It solves the difficulties of defect detection in PCB with high density and high integration and of inefficiency in real application.It possesses both theoretica l value and engineering applied value.