PERFORMANCE ANALYSIS OF NOISE FILTERS USING HISTOPATHOLOGICAL TISSUE IMAGES IN LUNG CANCER

Abstract

Digital Image Processing employing unique algorithms have found applications in many different fields. Noises associated with the histological images were needed to be removed for better image processing and diagnosis of cancer disease. Three different types of noise such as additive, random impulsive and multiplicative are normally associated with the any image. Here in the present work planned to remove the noise on medical cancer histological images for effective diagnosis of lung cancer. Since the type of noise is unknown, four basic filters viz., Average, Median, Gaussian as well as Wiener filter were analysed on the MATLAB platform. Further the filter performance through PSNR and MSE analysis. The intensity plot has also been drawn to describe the working of image filters. Overall filter experiments and its performance analysis conclude the median filter was effective in removal of the noise associated with the histological slides. The performance analysis was concluded that average median filter was effective for removing noise in biological sample.