Vision-based Inspection of PCB Soldering Defects

Vision-based inspection of printed circuit board (PCB) soldering defects was studied for preparing feature data and classifying the overall PCB on a prototype into different classes. The image developed using an sensor camera. Image augmentation conducted to enhance dataset volume. pre-processing included resizing, colour conversion, denoising. Watershed-based segmentation performed in post-processing segmented images; then, extraction curvelet transform prepare data. as statistical include kurtosis, contrast, energy, homogeneity, variance. These were analysed, percentage difference mean values between classes calculated. Kurtosis had highest among In comparison values, kurtosis obtained 4.97% class good medium condition; 17.02% bad 12.08% condition. Through this analysis, is considered more reliable machine-learning based classification project. extracted can be applied future studies classify solder joint by artificial neural network machine learning classification.