The Circuit Board On-board Components Detection System

Automation Optical Inspection (AOI) on printed circuit board assembly is an important research field in the electron industry product inspection. It aims at distinguishing components and checking their defects with high speed and accuracy on the high density circuit board, which adoptes the miniaturization component. It can save man labor and improve detection reliability, thus improves electron product mass and the efficacy.In this paper, AOI and its development is first studied, then a simple AOI system is developed.To solve the illumination problems, wavelet based homomorphic filtering is modified and used in our system. The wavelet coefficients in each level are filtered with a highpass filter to attenuate the low frequencies and enhance the high frequencies. Then the wavelet reconstruction is performed. Experimental results show that the new works well in AOI systems.To locate Mark points precisely and quickly, improved Hough transformation is adopted here to cut down points and dimensions in calculation, and thus label Mark points quickly. Experimental results illustrate the effectiveness of the proposed method.Since reports of components are obtained by CAD, feature database is produced by add a feature item ito the reports. The feature database contains the name, type, coordinate etc., and it can be easily used in components location, segmentation, recognition and defects detection.Finally components recognition and defects detection is performed.