Fuzzy Soft Thresholding Based Hybrid Denoising Model

This paper proposes a denoising model hybridized using wavelet and bilateral filters with fuzzy soft thresholding. The parameters of the proposed model are optimized with floating point genetic algorithm (FPGA). The model optimized with one image is used as a general denoising model for other images like Lena, Fetus, Ultrasound, Xray, Baboon, and Zelda. The performance of the proposed model is evaluated in denoising images injected with noises in different degrees; moderate, high and very high, and the results obtained are compared with those obtained with similar hybrid model with wavelet soft thresholding. Results demonstrate that the performance of the proposed model in terms of PSNR and IQI in denoising most of the images is far better than those with similar model with wavelet soft thresholding. It has also been observed that the hybrid model with wavelet soft thresholding fails to denoise images with very high degree of noises while the proposed model can still be capable of denoising.