Noise reduction of the real-time X-ray image based on modified adaptive local noise reduction filter

Real-time X-ray imaging technology is widely used in on-line weld defect detection in industry. In contrast to film images, real-time X-ray images have the characteristics of high noise and low contrast, which makes the detection of weld defects a difficult problem. Therefore, it is important to improve image quality by noise reduction algorithm. In this paper, the relationship between noise variance and grayscale value in real-time X-ray images is firstly analyzed. Then a modified Adaptive Local Noise Reduction filter is proposed to filter the image. Results show that the proposed noise reduction algorithm is much better than the original local noise reduction filter and the median filter, and avoid the confliction between the noise reduction and image details keeping of small defects.