FPGA based Hardback Conceptualization for Pothole Detection

Journal: GRENZE International Journal of Engineering and Technology
Authors: Roopa B S, R Manjunatha Prasad
Volume: 3 Issue: 3
Grenze ID: 01.GIJET.3.3.11 Pages: 20-24


Edge detection forms the primary path for any object detection which in turn is a capital part of image segmentation in the procedure of object detection, feature extraction and so on. Edge signals the extremity between overlapping or adjacent objects. An edge is also an extremity between an object and the backgroundd, hence by identifying the edges accurately it will be easier for the extraction of the information regarding measuring of area, perimeter and shape. In the direction of pothole detection few existing methods like ranging sensors of high maintenance or usage of acceleration data is expensive and also not very accurate in providing the database of the pothole. The pothole detection using image processing will help in alerting the driver and turning away from the same.The data of edge detection is very large, so the accomplishment of the high speed of image processing is a challenging task. Here various edge detection algorithmic rules are implemented on pothole images with many constraints and compared for narrowing down to the most efficient edge detector for pothole images. Field Programmable Gate Array (FPGA) can surmount this difficult task and it is an promising device to realize real-time collateral processing for vast amounts of image and video data. The proposed work pretense the exploitation of Prewitt edge detection algorithm on FPGA for pothole application.