Measuring scour level using image processing

Abstract

Scour monitoring is a process to measure the level of soil erosion at the bridge pillars. Currently, the monitoring and the interpretation is done manually. This work proposes an automatic scour monitoring system that is able to detect and measure the level of scour. The system uses image processing techniques such as image inpainting, Hough transform to detect the level of scour, and artificial neural network to measure the scour level and scale numbers. Results show that the scour level can be detected automatically for even and uneven soil, and the scour level can be measured automatically and accurately.