In the process of image acquisiting, the image quality will be degraded due to a variety of factors. Image enhanceing is aimed at highlighting some interested information that is easy to analyze for people and machine and inhibiting some useless information to enhance the image value. Image sharpening is a partion operation method of image enhancing in spatial domain, and its purpose is to enhance and judge the edge of the image and profile information and the specific method of the image sharpening uses differential to make the edge so prominent and clear.
The most commonly used method of image sharpening is gradient sharpening. But apart from the gradient algorithm, image sharpening methods also have Roberts, Prewitt, Sobel, Laplacian and etc. These methods were introduced, compared and analyzed. And it introduces the structure and characteristics of the DM642 DSP. Finally,a part of the image sharpening algorithm is achievd and the results afe recorded.
Through the simulation and comparison of the various algorithms, each algorithm has its own advantages and disadvantages. After the features of the image using in this paper are analyzed, it improves the Laplacian algorithm contrapositively, namely using high-elevating filtering to improve the brightness of the image. Experimental results show that the method is feasible and achieves the desired sharpening effect.