Automatic Detection System Of Circuit Board Defects Based On Image Surf Feature

With the rapid renewal of various electronic products,the defect detection of circuit board is becoming more and more important in the production process.At present,the production enterprises mostly adopt the combination of optical inspection instruments and manual inspection,which has the problems of high cost and incomplete automatic inspection process.In recent years,with the rapid development of computer image processing,defect detection of PCB Based on machine vision has become a research hotspot.In this paper,according to the actual needs of a PCB manufacturer,an automatic defect detection method based on image surf feature is proposed,which has a certain practical application value.The specific contents include:In this paper,the camera is used to obtain the circuit board image placed at any angle under the illumination of ordinary fluorescent lamp.Aiming at the noise produced in the process of image acquisition and the deviation of the circuit board position in the image,the Gaussian filter,median filter and other methods are used to remove the noise.The Hough detection method is used to correct the angle tilt of the image,and according to the different background of the circuit board Two segmentation methods are implemented and a better image preprocessing effect is obtained.In this paper,surf feature descriptors are used to extract and match the image features,and RANSAC algorithm is used to correct the matching errors.Finally,the 3 * 3 affine transformation matrix is obtained,which makes the effective part of the circuit board image consistent after affine transformation.Finally,the standard image and the detection image are stretched through gray scale,and the image is subtracted and binary after gamma transformation and corrosion expansion Then,the binary part is further corroded and expanded to eliminate small errors,and the defective part of the circuit board is obtained.In order to verify the effectiveness of the defect detection method proposed in this paper,the Open CV open source visual library is used to realize the engineering encapsulation of the whole detection in Python language environment,and the system operation from photographing to defect detection is realized,and the accuracy andreal-time performance of the system are tested and evaluated.The results show that the fast detection of PCB defects based on image proposed in this paper can achieve good detection results in a short time.