|
چکیده
|
Digital X-ray radiography is a well established method for non-destructive detection of corrosion and measurement of wall thickness of pipes. Due mainly to the unavoidable detection of scattered X-rays as well as the ever-present system electronic noise, the quality of the acquired radiographs is often low; the quality can however be enhanced by applying various image processing methods. In this research, Wavelet Transform (WT), Gabor Filters (GF) and Shearlet Transform (ST) have been utilized in a background removing algorithm to achieve the required radiographic image quality improvement. The algorithm first decomposes the image into different levels using the WT, GF or the ST method, and the automatic thresholding level, which is based on Rayleigh distribution, is then applied. The output images (with the background removed) were then reconstructed by subtracting the modified component from the original radiograph. The results of the study showed that the reconstructed images yield significantly greater contrast than the original radiograph, the wall thickness measurements have higher precision and defect detection has greater specificity for defect in all regions. Also, the outcomes show that the algorithm has high analytical efficiency and is capable of accurately measuring pipe wall thicknesses and corrosion regions.
|