Pulmonary CT image denoising algorithm based on curvelet transform criterion

Abstract

Noise pollution on pulmonary CT images is always unavoidable during the acqisition of the images, traditional denoising algorithm can't successfully get rid of noise on pulmonary CT images effectively without destroying the texture and edge features. In order to filter noise of pulmonary CT images and keep the edge and texture signal in images, this paper presents a pulmonary CT images denoising algorithm based on Curvelet transform. First, algorithm carries on Curvelet transform to noisy pulmonary CT images. Then algorithm constructs direction criteria and scale criteria for noise determination in the curve wave domain, and noisy pulmonary CT images is denoised. At last, curved wave inverse transformation is performed and the denoised pulmonary CT images is obtained. Experimental results show that this method has better effect for keeping edge and visual smooth, compared with traditional curvelet threshold shrink method.