Artificial Haze Immune Algorithm for Image Processing

Abstract

The Intelligent Transportation Systems bring a safely and comfortable motorized society, which are based on image processing such as predicting/detecting the danger of vehicles collecting the transport information to control the traffic flow on traffic control systems etc. However, with the pollution of environment, the fog/haze becomes a serious problem, causing the image deterioration or degradation and the Intelligent Transportation Systems lose their functions. This paper proposed an immunological method of image processing for detecting the fog/haze in the image to support Intelligent Transportation Systems. This state-of-the-art method is based on the theory of biological defence system, which unites the reducing of fog/haze and edge detection. The experimental results show that our proposed haze immunized algorithm can remove effect of haze on image processing algorithm. Compared to conventional de-haze image processing algorithm, our proposed algorithm unitizes bio-inspired algorithm to achieve efficient hardware consumption. The results of FPGA implementation show less hardware usage than conventional method.