Saliency-Guided Deep Learning for Bridge Defect Detection in Drone Imagery

TL;DR

This paper introduces a novel drone-based method combining saliency detection and YOLOX deep learning to automatically identify and classify bridge defects efficiently, promising enhanced accuracy and speed for infrastructure inspection.

Contribution

The paper presents a new two-stage framework integrating saliency-guided defect proposal and YOLOX detection on saliency-enhanced images for bridge defect detection.

Findings

High accuracy in defect detection on standard datasets

Improved computational efficiency over existing methods

Potential for deployment in autonomous inspection systems

Abstract

Anomaly object detection and classification are one of the main challenging tasks in computer vision and pattern recognition. In this paper, we propose a new method to automatically detect, localize and classify defects in concrete bridge structures using drone imagery. This framework is constituted of two main stages. The first stage uses saliency for defect region proposals where defects often exhibit local discontinuities in the normal surface patterns with regard to their surrounding. The second stage employs a YOLOX-based deep learning detector that operates on saliency-enhanced images obtained by applying bounding-box level brightness augmentation to salient defect regions. Experimental results on standard datasets confirm the performance of our framework and its suitability in terms of accuracy and computational efficiency, which give a huge potential to be implemented in a self-powered inspection system.