A Fault Detection Method for Tension Clamp Compression in Transmission Lines Based on AKAZE and APAP Image Correction

this study proposes an intelligent detection method for the compression quality inspection of tension clamps in transmission lines, addressing the issues of low efficiency and insufficient accuracy in current methods. Firstly, X-DR image quality is enhanced through a triple preprocessing process, including Gaussian filtering, histogram equalization, and gamma correction. Secondly, the APAP algorithm is employed to construct a local homography grid model, effectively addressing parallax distortion in multi-view image stitching. Subsequently, the AKAZE algorithm is used to extract nonlinear scale-space features and establish a defect recognition model. Experimental results demonstrate that the proposed method accurately detects typical compression defects such as under-pressure, over-pressure, and pressure loss, with significantly higher detection accuracy compared to traditional methods. This research provides a novel technical approach for the quality inspection of critical components in transmission lines and holds significant engineering application value.