Nowadays, millions of people die each year in traffic accidents. Intelligent transport systems (ITS) integrated electronics, communications, control and sensing technologies, to enhance the efficiency of the transport system and the running safety of human-vehicle. Major automakers in Europe, United States and Japan have invested abundant resources in the development of the automobile safety assistant driving system. In many relative studies researchers found that traffic accidents can be broadly divided into ??human?? and ??weather?? caused, while the natural factor ??weather?? is more difficult to be tackled. It was not resistant such as sandstorms, snowstorms, fog, rain, etc. Such situation leads to more difficult driving and more accidents. In the past, most use of the camera lens to get images, and rare usage of laser or infrared due to cost considerations. Hence, we get an image for further evaluation and analysis, via applying good pre-processing method, to solve such problems caused by weather factors. We have designed a driving assistance system which integrates specific functionalities of lane detection, departure warning, vehicle distance measurement, collision warning and to deal with the vehicles in the neighboring lane and stabilize the image sequence of the captured video. The conducted system is implemented by a camera mounted inside the vehicle and image processing technologies. As the adequate warning messages are granted, drivers can generally have enough time to response. We will use a novel interactively recurrent self-evolving fuzzy CMAC to solve the image problems caused by weather (such as fog), acquire a clearer image and improve the corresponding accuracy of the lane detection, departure warning, vehicle distance measurement, collision warning and to deal with the vehicles in the neighboring lane and stabilize the image sequence of the captured video.. Finally, we will be integrated to form a completed system.