Implementation of Adaptive Thresholding Method for Road Surface Damage Detection in Digital Images

Abstract

Road surface damage such as cracks and potholes can disrupt comfort and compromise the safety of road users. Therefore, a method is needed to assist in the automatic identification of road damage. This study applies digital image processing techniques to detect road damage using images captured by a smartphone camera. The image processing procedure consists of several stages, including converting color images to grayscale, reducing noise using Gaussian Blur, segmenting images using the Adaptive Thresholding method, and detecting damaged areas using contour detection. The system then identifies the type of damage based on the shape of the detected objects and classifies road conditions into three categories: smooth, cracked, and pothole. The results of this study are expected to help identify road damage more quickly and objectively through digital image processing.

Keywords: Digital Image Processing, Adaptive Thresholding, Contour Detection, Road Damage, Smartphone