Published Paper Details:

Image data captured from outdoor visual devices are usually degraded by some external factors such as fog, rainfall, snow, smoke and haze. Hence the weather condition plays a prominent role in disturbing the public transportation systems as well as some advance assistance systems which try to capture the data from outdoor cameras and then function the system.Hence in order to avoid this disruption of images from outdoor cameras, one should able to identify the weather situation and what are the possible factors that may cause due to current weather images.In primitive days there is no such automatic method which can identify the weather images and process those images and find out the possible factors related to that weather image. Hence this motivated me to develop this proposed model, in which we try to collect some weather images and apply image processing techniques and CNN model to identify the weather image and its possible outcomes in order to take preventive steps from deweathering operations (e.g., removals of haze, rain, or snow) would be correctly triggered accordingly. Here we try to collect some weather images from jehanbhathena/weather-dataset and then apply VGG-19 as best CNN model to train the application and our experimental and simulation results clearly state that our proposed system achieved nearly 96 % of accuracy compared with other primitive models and hence this proposed VGG-19 is best and accurate in order to predict the weather possibility from weather images.