Sectional image restoration of sintering machine tail based on dark primary prior

The sintering tail section images collected relying on the machine vision technology, then through digital image processing, can effectively reflect the pros and cons of the production of sinter quality. But the sintering machine tail part working conditions are very bad, have the massive mist and dust, causing the sintering machine tail section image which gathered in aspects of the brightness, the contrast gradient and the clarity have the very tremendous influence, which causes the image degenerating and brings difficulties to image characteristic extraction and pattern recognition. Based on image degeneration mode in the dust environment, uses one kind of method to remove the image dust influence effectively. First the restore formula and the direct transmission capacity formula are derived through the degenerated model and the dark channel prior. Next acquires the dark primary color of image through the dark channel prior knowledge, and estimates the atmospheric light ingredient. Then calculates the depth map according to atmospheric light ingredient and the direct transmission capacity formula which are estimated to. Finally restores the dust image based on the depth map and the restore formula. Simulations in the Matlab platform demonstrate this algorithm can efficiently improve degeneration phenomenon of sintering machine tail section image and enhance the clarity of image.