Visibility enhancement of scene images degraded by foggy weather conditions with deep neural networks
F.Hussian and J.Jeong, Journal of Sensors,pp.1-9,2016
This paper focuse on removing fog that would be helpful for the drivers to avoid collision. Foggy weather conditions have lower contrast and faded scene visibility making it tuff for the drivers. This can be removed by employing deep neural networks. The NN consists of input layer,output layer and many hidden layers, it uses backpropagation algorithm with iterative gradient descent and hyperbolic tangent as the activation function. The working is same like regular NN.