Wavelet based fusion of fuzzy enhanced images captured under improper illumination

摘要

To increase the application of images captured under improper illumination, preprocessing of underexposed and overexposed images before their fusion in wavelet domain is proposed. Since images contain some fuzziness in nature, hence both the underexposed and overexposed images are preprocessed using fuzzy enhancement method based on Gaussian intensification function. These enhanced images are then decomposed using Á-trous algorithm in wavelet domain where the Á-trous coefficients are fused using average based fusion rule. Finally, the inverse Á-trous algorithm reconstructs the fused image that has proper contrast and contains all the detail present in input images. Both subjective and objective assessment is done to evaluate the performance of the proposed algorithm. The fused image obtained by fusing fuzzy enhanced overexposed and underexposed images using Á-trous algorithm is better than those achieved by fusing images using DWT with fuzzy enhancement, DWT and averaging method.