Abstract: In electrical engineering experimental teaching, component quality directly affects experimental results and circuit safety. To address the issues of low efficiency and strong subjectivity in manual inspection, this paper proposes a machine vision detection algorithm based on image feature matching. The algorithm preprocesses images through adaptive grayscale correction and mean filtering, extracts geometric features such as contour area and perimeter using Otsu’s threshold segmentation method, and constructs a weighted feature similarity model to automatically identify defects including scratches, breakage, deformation, and missing pins. Theoretical analysis shows that the algorithm exhibits strong robustness and outperforms traditional single-threshold methods, providing theoretical support for intelligent component inspection.