Published Paper Details:

The intent of this paper is to establish a model to classify outdoor images to different weather classes and then using this classification as basis of many actions that can be provided to driverless cars. While performing the literature survey about field related to weather prediction we found out that the best results for weather classification are obtained while using the CNN models. This paper proposes a method of implementing convolutional neural networks to get the desired result of classifying separate weather conditions into four classes, namely foggy, rainy, sunny and snowy. And according to the weather condition predicted, the project will provide the necessary actions that need to be taken by the driverless cars. Convolutional neural networks (CNNs) are widely accepted and acknowledged in the fields of image analysis, image recognition, video recognition, face recognition, image classification and activity detection. CNNs have images as their input; process those images; assign adaptive weights and biases to many features of the image; and then classify them into various categories.