Abstract: Image recognition technology is one of the typical application scenarios of artificial intelligence in the field of weld seam radiographic testing. Conducting research and application of image recognition technology in industrial weld seam detection and intelligent monitoring is of great significance for promoting the intelligent development of non-destructive testing. The preprocessing of defect images in radiographic testing can simplify complex images in a short period of time, laying a solid foundation for subsequent defect recognition. Due to the narrow gray range, low contrast, and high noise in the original X-ray detection image, different denoising and contrast enhancement image preprocessing methods were used to solve this problem. X-ray film preprocessing experiments were conducted, and parameters were optimized and algorithms were improved based on actual detection results. The experimental results showed that in terms of noise reduction, the median Gaussian combination filter had a better noise reduction effect; In terms of contrast enhancement, linear transformation had a better effect on contrast enhancement.