Scholarlink

<p>Chipping is a common surface defect in automotive coatings, caused by the high-velocity impact of tiny stones and gravels that leads to localized coating detachment and loss. Conventional methods for evaluating chipping resistance typically rely on visual inspection according to special standards, which is subjective and prone to inconsistency. This study aims to develop an automated machine vision and image processing method for the objective detection and grading this effect. Coated panels were prepared and tested in accordance with the PSA D241312-H (Peugeot) standard to report varying degrees of chipping. High-resolution images were acquired under uniform lighting. An image processing pipeline combining contrast enhancement, thresholding, and contour-based analysis was implemented to identify and classify damaged areas based on the chip size and chip distribution. The results demonstrated that the proposed system could accurately detect even subtle or complex defects that were often challenging for human observers, while providing consistent and reproducible grading. In contrast, visual evaluations by four experienced inspectors showed substantial variability, as reflected by large standard deviations across samples. Despite this disagreement, the system achieved a strong correlation with the inspectors’ mean ratings (R = 0.86), highlighting both its accuracy and its ability to provide a more stable reference. Moreover, the automated approach reduced inspection time by approximately a factor of seven compared to manual evaluation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A new approach to detect and grade the chipping defect of automotive coatings using a machine vision system

  • 期刊论文

  • 发表日期：

    2026年5月19日

查看全文
  • Kazhal Afshargoli,
  • Saeideh Gorji Kandi,
  • Mohsen Mohseni,
  • Hossein Yahyaei

摘要

Chipping is a common surface defect in automotive coatings, caused by the high-velocity impact of tiny stones and gravels that leads to localized coating detachment and loss. Conventional methods for evaluating chipping resistance typically rely on visual inspection according to special standards, which is subjective and prone to inconsistency. This study aims to develop an automated machine vision and image processing method for the objective detection and grading this effect. Coated panels were prepared and tested in accordance with the PSA D241312-H (Peugeot) standard to report varying degrees of chipping. High-resolution images were acquired under uniform lighting. An image processing pipeline combining contrast enhancement, thresholding, and contour-based analysis was implemented to identify and classify damaged areas based on the chip size and chip distribution. The results demonstrated that the proposed system could accurately detect even subtle or complex defects that were often challenging for human observers, while providing consistent and reproducible grading. In contrast, visual evaluations by four experienced inspectors showed substantial variability, as reflected by large standard deviations across samples. Despite this disagreement, the system achieved a strong correlation with the inspectors’ mean ratings (R = 0.86), highlighting both its accuracy and its ability to provide a more stable reference. Moreover, the automated approach reduced inspection time by approximately a factor of seven compared to manual evaluation.