Road sign detection and recognition in panoramic images to generate navigational maps

The paper proposes an approach to recognize road signs in Google Street View panoramic images. The color images formed by special queries from the Google site with a 640×640 resolution serve as the input data for recognition. The recognition of the road signs is carried out by the use of the specific, while the classification is performed by the use of the neural network. The coordinates of the road signs in the image help you get the latitude and longitude of the sign location. The navigation map is made up with the coordinates of the road signs location and their type.