U<sup>2</sup>D<sup>2</sup>PCB: Uncertainty-Aware Unsupervised Defect Detection on PCB Images Using Reconstructive and Discriminative Models
Chen Chang-lin, Qiman Wu, Jin Zhang, Haojie Xia, P. Lin, Yong Wang, Mengke Tian, Rencheng Song
Abstract
The defect detection of printed circuit board (PCB) images face challenges such as limited sample number, imbalanced sample types, and varying detection reliability. To address these issues, this paper proposes an uncertainty-aware unsupervised detection model on PCB images, short for U <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> PCB. The proposed method utilizes two U-Net networks to serve as the reconstructive sub-network and the discriminative sub-network, respectively. The former one reconstructs defect-free PCB images from defective PCB images, while the latter segments the defects and evaluates the defects uncertainty with the concatenated inputs of the defective and reconstructed images. The U <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> PCB model is trained in an unsupervised manner with only defectfree images embedding with multi-scale artificial defects. Experimental results on the public PCB defect dataset and DeepPCB dataset demonstrate the effectiveness of the proposed method. The mean average precision (mAP) is 99.29% on the PCB defect dataset, while it reaches 95.78% on the DeepPCB dataset. These results are competitive to those of state-of-the-art fully supervised methods. The findings of U <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> PCB highlight the potential significance of employing unsupervised learning techniques for PCB defect detection.