Impulse Noise Removal Using Adaptive Bilateral Filter with Robust Noise Detector

Impulse noise removal is important as the images may get corrupted during the acquisition, transmission or storage. A variety of noise removal algorithms are reported in the literature to denoise the image corrupted by impulse noise which may be random valued or can have a minimum and maximum value of gray scale. We propose few denoising techniques which are more suitable for removing impulse noises from the corrupted image. Salt and pepper noise, random valued impulse noise and mixed noises are considered in this paper. To improve the efficiency and robustness of denoising, we initially estimate the noise pixels in the image and further it can be removed with robust filtering technique. Also, the kernel shape and the size is determined by the noise distribution and the density. The simulation results show that the proposed filter can be utilized for image with high noise density and it can be easily implemented in the hardware to make it usable in the real-time environment.