Accuracy of Computer Vision Hough Line Transform for Water Level Measurement

Abstract

Floods are natural disasters that often occur in countries with high rainfall, such as Indonesia. Floods occur due to sudden changes in water level; therefore, the dam cannot withstand the increase in water causing water to overflow into the riverbank area. As a result of this issue, a system was created that can be used to provide information about river water levels. This system relies on computer vision to program the camera to monitor the water level through image processing. Image processing allows the camera to process the image of the water surface detected at the ditch’s edge. The edge detection result is used as a water level detector using the Hough line transform, which generates a line to indicate the water level. The height detected by image processing differs from the actual height on the detected surface. The minimum and maximum error reading for both methods recorded in this experiment is 0,8 % and 16,67 %, respectively