Consultancy

Abstract

It is important for engineers to know the road conditions for smooth conduct of traffic. Civil Engineers must spend lot of efforts in knowing the road cracks and their size, shape and location co-ordinates. With the spread of digital camera, GPS tracking systems and image processing, this problem of road crack detection can be solved with help of Artificial Intelligence and Image Processing. In this work a convolution neural network has been developed which can identify the road crack in an image captured by a camera on a vehicle. This CNN module served as server side part for detection of road crack. Another part of model, serving as client side, was affixed to the moving vehicle for sending images to the server along with location co-ordinates. Thus a client-server based model is developed in this research which can track the road condition and also send SMS to authority, from the server side, for repair of the road.

Project Model

The project model had a client server architecture. The client was developed using Arduino edge device which had a GPS sensor. The server was made of image processing software which can receive road images and test using CNN model if the image has a road crack or not. The details of model are pictorially given in Fig. The modules of model are described as below: