<p>Tension clamps, which are integral components of transmission lines, require regular inspection for defects. Compared with conventional approaches, the drone equipped with X-ray inspection equipment provides a better solution for evaluating the internal state of tension clamps. However, the motion-blur resulting from vibrations of drone can adversely affect subsequent defect detection, highlighting the importance of deblurring processes. In this paper, we present a methodology for motion deblurring of X-ray images of the tension clamps. Initially, the Object Detection Network is employed to discern and isolate regions corresponding to tension clamps. Subsequently, we introduce the conditional cropping strategy to improve the network’s ability to extract blur information from more valid parts of the image. In terms of network module improvement, we introduce the DSKconv/DSCKconv to significantly enhance the network’s capability to comprehend image content. Our deblurring methodology attains PSNR and SSIM of 29.4071 and 0.8514 respectively in deblurring quality, and can achieve 57.667G FLOPs and 18.968M Params in terms of computational efficiency. Corresponding datasets are created for both the Object Detection Network and the Deblurring Networks. We provide an excellent motion deblurring solution for X-ray images of tension clamps taken by drones.</p>