A new method for foggy image enhancment

As a common natural phenomenon, fog makes photographed images blurred, thus bringing great difficulty for monitoring of outdoors image, target identification and tracking and so on. In this paper, the authors propose a new method for foggy image enhancement that integrates multilevel wavelet decomposition, the auto-adapted LUM filter, soft threshold and so on. Firstly, carry on the multilevel wavelet decomposition to the image, and then obtain the low-frequency component and high-frequency components of image, use the auto-adapted LUM filter to low-frequency component, while utilize soft threshold based on Bayes estimation to process high-frequency components, and eventually carry on wavelet restructuring to the processed components. Through simulation, this method is proved to be superior to traditional methods in foggy image enchantment.