A Method of Crack Defects Real-Time Image Enhancement in Continuous Casting Billet

Abstract

In the embedded image detection of continuous casting billet defects, the image enhancement of crack is an important criterion in the subsequent recognition. The effect of image enhancement, especially the real-time of image enhancement, decides whether the system can be adopted or not. Based on the dual-core OMAP3530, we set up an embedded real-time online testing hardware device. Taking the system's real-time and the environmental noise of the high-temperature scene into account, the billet image, obtained by the CCD sensor, is simply and effectively enhanced by going through the Laplacian sharpening and the Histogram equalization. Matlab simulation analysis improves the Contrast Improvement Index (CII) of image and cuts down the simulating time; and it provides a powerful precondition for the subsequently billet defects real-time identification by DSP hardware.