FFBF: cluster-based Fuzzy Firefly Bayes Filter for noise identification and removal from grayscale images

Image denoising gains more attention in the field of image processing, which is essential to sustain the originality of the digital images in order to preserve all the essential information buried in the image. Even though lots of denoising techniques are available, the existing methods failed to denoise the image efficiently, and they are applicable only with lower noise probability. Thus, this paper proposes a Fuzzy Firefly Bayes Filter (FFBF) to perform the noise identification and removal. FFBF employs the Ck-based firefly Bayes algorithm and probabilistic clustering for identifying the presence of noisy pixel in the input image. The Ck-based Firefly Bayes algorithm is newly proposed by integrating the cuckoo search optimization, firefly optimization, and Bayes Classifier and it is based on the maximum posterior probability objective function. The proposed algorithm provides the best solution for the formulation of the binary matrix using the Bayes Classifier, which is subjected to fuzzy-based image denoising. The paper uses two standard images for experimentation, and the comparative analysis is performed in order to determine the superiority of the proposed method. The PSNR, SSIM, and SDME obtained for the proposed method are greater when compared with the existing methods, and the proposed method attained a maximum PSNR, SSIM, and SDME of 45.1696 dB, 0.8260, and 59.9684 dB.