Image Enhancement Methods Research In The Heavy Fog

Under the condition of fog weather, the image captured by outdoor monitoring system has low contrast because of the scattering effect of atmosphere, and there is certain effect on human production and life. Consequently, it is significant to study on image clearness technique in misty weather.At present, clearness technique of fog image is mainly sorted into the image restoration method based on model and the image enhance method based on non-model. the point of view of image enhancement was studied in this paper, the following parts were the main content:First, the research actualities of digital image enhancement technique was summed up in detail, then, based on the conventional spatial domain and frequency domain enhancement algorithm, its achievement on the field of image enhancement was also summarized.Second, we introduced the developed history of fuzzy set theory, and its application in image processing, demonstrated the basic theory of fuzzy enhancement, analyzed how Pal applied to image enhancement, and then pointed out its disadvantages in image enhancement, concluded some improving algorithm that other researchers had studied on its basis, meanwhile two new methods were proposed according to the fuzzy enhancement of function and fuzzy enhancement operator:one improving overall fuzzy enhancement algorithm, the other improving partial fuzzy contrast enhancement algorithm.Finally, as to the problem that conventional image enhancement algorithm increased image noise at the time it enhanced image information, this coupled contradiction that conventional image information and noise enhanced simultaneously can be solved effectively by wavelet multiscale distinguished analysis. For the main contour of image mostly embodied in low-frequency part and the section of detail and noise high-frequency part, as a result, we availably separated image approximate information from detail and noise by using technology based wavelet multiscale transform, enhanced the low-frequency decomposition coefficient obtained after transforming, and then the main noise information can be filtered by adopting threshold de-noising method to the high-frequency decomposition coefficient, which well eliminated the influence of noise on image enhancement. The enhancement algorithm based on theory of fuzzy set and wavelet transform is a considerably novel method, the additional noise can be filtered while enhances image information, and considerably fair experimental result can be achieved in terms of enhancing image contrast.