Research On Visibility Detection Algorithm For Fog Image

Visibility is closely related to fog,haze,rain and other weather.Low visibility not only makes the outside world blurred,but also brings adverse effects on people’s production and life,especially in transportation,electricity supply,agricultural production and other fields.With the continuous economic growth in recent years,the annual incidence of accidents on land,sea and air transport is increasing.The main reason is that the low visibility caused by the foggy environment.Therefore,studying the distance of visibility in foggy weather has great practical significance to people’s production and life.Firstly,the paper introduces the existing visibility detection algorithm and analyzes its advantages and disadvantages.Then,based on the analysis of the relevant feature information of fog images and fog-free images,the paper proposes a new method of visibility detection algorithm based on the dark channel prior.This method combines the prior principle of the dark channel with the parallax characteristics of the binocular vision,extracts the transmittance and depth of field to calculate visibility.The experimental results show that the algorithm can effectively detect the visibility distance of fog and the robustness of the algorithm is better.In order to overcome the complex operation of the algorithm,this paper also proposes a visibility detection method based on a single image of SSR.The method is based on the retinex theory and Lambert-Beer law which describes the attenuation of atmospheric luminous flux in the air.Thus,by using the relationship between the transmittance and the luminance component in the irradiated image,the visibility of the fog image can be estimated.The algorithm simplifies the computational complexity of visibility measurement to a certain extent,improves the operating efficiency of the algorithm,and at the same time satisfies the accuracy of the detection.Finally,this article takes visibility as a new quality assessment index to evaluate fog image sharpness.According to experimental data,visibility can more directly reflect the effect of image sharpening.