Pixel Fusion Based Curvelets and Wavelets Denoise Algorithm

—Curvelets denoise approach has been widely used in many fields for its ability to obtain high quality result images. However artifacts those appear in the result images of curvelets approach prevent its application in some fields such as medical image. A denoise algorithm is proposed to perform pixel fusion to result images of curvelets and wavelets approaches. Firstly curvelets denoise result image and hidden Markov tree based wavelets denoise result image are calculated. Then image regions are analyzed with quadtree decomposition. Finally result images are obtained with weighted pixel fusion method. Experimental results demonstrate that the algorithm improves the visual quality of result images efficiently and suppresses the artifacts in result images evidently. And the results of the algorithm applied to ultrasonic medical images also indicate that the algorithm can be used efficiently in medical image fields.

• Type: Journal
• Year: 2007
• Where: ENGL
• Authors: Liyong Ma, Jiachen Ma, Yi Shen
• Keywords: ENGL 2007, Image, Medical Image, Pixel Fusion