Bayesian Outdoor Defect Detection

TL;DR

This paper presents a Bayesian defect detection method tailored for motion blurred images on rough surfaces, utilizing reflected non-local priors to improve accuracy in distinguishing defects from non-defects.

Contribution

It introduces a novel Bayesian detector with reflected non-local priors that effectively suppress non-defect pixels, enhancing defect detection accuracy on motion blurred images.

Findings

Superior performance in eliminating non-defect points

Successful identification of hail damages on drone images

Enhanced overall defect detection accuracy

Abstract

We introduce a Bayesian defect detector to facilitate the defect detection on the motion blurred images on rough texture surfaces. To enhance the accuracy of Bayesian detection on removing non-defect pixels, we develop a class of reflected non-local prior distributions, which is constructed by using the mode of a distribution to subtract its density. The reflected non-local priors forces the Bayesian detector to approach 0 at the non-defect locations. We conduct experiments studies to demonstrate the superior performance of the Bayesian detector in eliminating the non-defect points. We implement the Bayesian detector in the motion blurred drone images, in which the detector successfully identifies the hail damages on the rough surface and substantially enhances the accuracy of the entire defect detection pipeline.