Car Eye

Abstract

Autonomous cars are frequently provided with a camera that records the environment in front of the vehicle. This camera's ability to recognize weather conditions can help in the improvement of many applications as well as the development of new ones. The aforementioned adverse weather conditions, which we are concerned about within our project are rain, fog, and bad illumination at night. These conditions negatively affect the quality of the stream coming from the camera into the system inside the vehicle. We want to improve the quality of this stream and keep it enhanced at all times. This can be done by doing the following: determining the existing status of the weather and then improving it through deep learning techniques. The proposed system will be designed based mainly on computer vision and image processing, using deep learning techniques (neural networks like CNN), we will use computer vision to classify the weather and image processing to make enhancements. Our Project is a software solution that works within an autonomous vehicle to enhance the current visual stream; the project has two main modules, the first one is classification module and the second one is enhancement module, the first module will classify the state of the current weather (rain, fog and night) then the labels of the current weather will be passed to the second module that enhances the visual stream so, you can have a super clear vision to the surrounding