<p>High-quality inspection of printed circuit boards (PCBs) is crucial due to the increasing complexity and prevalence of electronic devices. X-ray imaging, a commonly used non-destructive testing method for detecting soldering defects in PCBs, traditionally relies on cumbersome and inefficient human evaluation. Furthermore, the imaging noise introduced by the development of low-dose X-ray imaging technology increases the difficulty of separating defect features from background information. This study innovatively utilizes stochastic resonance in nonlinear systems, actively leveraging the constructive effects of noise to extract anomalous features from X-ray inspection images, thereby significantly enhancing detection accuracy and efficiency. By analyzing the dynamic behavior of equilibrium points and attractor trajectories in bistable and tristable systems, we elucidated the operational mechanisms of logical stochastic resonance. Based on this, a minimal nonlinear array element capable of achieving XOR logic stochastic resonance was constructed, enabling a novel automatic method for extracting soldering defect features in PCBs. Experimental results demonstrate that this method exhibits higher accuracy in detecting various defects, significantly outperforms traditional differential morphology methods, and shows clear advantages over deep learning models that rely on large-scale labeled data in terms of computational efficiency and model interpretability. This method provides an effective solution for the efficient batch detection of PCB soldering quality.</p>