Edge Detection Algorithm And Its Application In Image Noise Reduction

In the field of image processing, the research of edge detection and noise reduction plays an important role in the theory and practice. Edge detection is one of the basic contents on image processing. The results of edge detection provide the required information for noise reduction. So the technology of edge detection is often used as the basis of image noise detection and plays an important role in the field of image noise reduction.Image edge detection provides important basic information for image segmentation and noise reduction. At present, the traditional edge detection algorithm includes Canny operator, Soble operator, LOG operator and so on. On the other hand, the edge detection algorithms combined with new technology get a fast development. The edge detector based on fuzzy inference rules is a new direction of research and the detector has a good anti-noise effect for Salt and pepper noise. But it is sensitive for the gaussian noise. When the gaussian noise is strong, the continuity of edge points is destroyed and is judged as noise points. The wrong judgement of edge points cause the results of edge detection to fail. In this paper, an improved image fuzzy edge detection algorithm is proposed to solve this problem. This algorithm can adjust noise edge detection scheme according to the noise intensity and fuzzy entropy. It uses fuzzy inference rules detection algorithm when the noise is weak and uses the improved fuzzy inference rules detection algorithm when the noise is strong. The improved algorithm combines fuzzy denoise algorithm and fuzzy inference rules detection algorithm to improve the effect of anti-noise.The results of image edge detection are often used as adjustment parameter for sub-regional processing of image denoise. At present, the image noise reduction algorithm can be divided into local noise reduction algorithm and nonlocal noise reduction algorithm according to filtering scope. The non-local means filter is an efficient denoising algorithm and attracts a wide spread attention. However, the filtering result produces artifact that affects the filtering image quality. Owing to the shortcomings of existing methods, an adaptive algorithm based on edge detection is presented in this paper. Firstly, the algorithm realizes the automatic tuning of filtering coefficient. Secondly, adjustment parameter is used to combine non-local means algorithm and isotropic algorithm. The adjustment parameter can be obtained by Pal-king algorithm, local variance algorithm, fuzzy inference rules detection algorithm and improved fuzzy inference rules detection algorithm. Finally, the selected adjustment parameters are used as adjustable factor to combine non-local means algorithm and isotropic algorithm for better effect of noise reduction.We verify the effects of improved fuzzy inference rules detection algorithm and adaptive NLM algorithm based on edge detection through experiments. Experimental results show that proposed algorithms achieve better effect.