Obstacle Detection on Road and Indication on Map at Rainy Times

The majority of accidents that occur during rainy seasons are caused by poor road conditions. The aim of this effort is to detect road cracks, potholes, and drainage holes that are covered by water. The detected information is stored in the database. At rainy times we fetch those obstacles on the path using efficient algorithms and display them on the map so that the users are informed about the road conditions and the potholes ahead and they can lower their speed and take necessary actions to avoid accidents. The camera is used to capture the road scenes. The gathered road images and position information were processed by a convolution neural network classifier that used the Keras sequential model with the Adam optimizer.The classified obstacles and location details are transmitted to the server and stored in the database securely.