Defect Edge Detection in Blockboard X-ray Images by Shannon Entropy

Abstract

Abstract A Shannon entropy-based image processing approach is introduced and applied to the blockboard X-ray images obtained from nondestructive scanning. X-ray nondestructive testing technology has been applied to the detection of internal defects in blockboard. In this paper, we select a probability distribution function to calculate the Shannon entropy in images processing. And we define a novel Defect Edge Index (DEI) for analyzing the defect edges in images. Through studying and processing the DEIs, the defect edges extracting is achieved. Furthermore, a new index E for quality evaluation was also studied out by the DEIs. The number of E can demonstrate the quality of examined blockboard. Experimental results display that the image processing method based on Shannon entropy theory is effective and the quality evaluation index E is accurate. Thus, a promising method for detecting and analyzing defect edges in blockboard X-ray images is provided.