Implementation of noise and hair removals from dermoscopy images using hybrid Gaussian filter

AbstractSkin cancer is the leading type of cancer which causes millions of deaths of human beings. Early identification and appropriate medications for new harmful skin malignancy cases are fundamental to guarantee a low death rate as the survival rate. To effectively classify skin cancers, it is necessary to remove the noise and hair from dermoscopy images. But the existing methods failed to achieve maximum efficiency, to solve these problems. In our proposed system we implemented a multi-resolution joint Gaussian filter (GF) with a hybrid decision-based window mask generation (HDWMG) approach for denoising of dermoscopy images. After performing the filtering operation, unwanted regions will be effectively detected by using the self-adaptive thresholding operation. The connected component label (CCL) method was applied to remove all unwanted objects present in the dermoscopy images. Then morphological area opening operation is used to identify the hair threshold levels and finally fast marching in-painting procedure is applied to remove the hair and restores the skin respectively. This work is implemented in Matlab platform, and extensive simulation results show that the proposed HDWMG-based GF performs superior in terms of both subjective and objective evaluation compared to state of art approaches. Further, it also tested for an image denoising application and shown that it enhances the noisy images without degrading the actual quality of the image.
Source: Network Modeling Analysis in Health Informatics and Bioinformatics - Category: Bioinformatics Source Type: research