Abstract: Due to the complexity of a highly dynamic X-ray imaging system and the complex structure of the workpiece itself, X-ray images often exhibit low contrast or unclear feature information. In this work, an image enhancement algorithm was proposed based on gradient field and non-local means to improve visual quality. A contrast adaptive enhancement model was constructed based on gradient field and local variance to improve the contrast of X-ray images; Then, the image was denoised using non-local means.Specifically, this paper constructed a non-local means algorithm based on Poisson distribution for image Poisson denoising, thereby obtaining an enhanced contrast field; Finally, an energy function was established and solved by using the gradient descent method to obtain higher quality images with small details.Experiments with two typical complex workpieces were performed, and the results verified the effectiveness of the proposed approach for image enhancement and defect detection.