Scholarlink

<p>In contemporary industrial production, accurately identifying steel plate surface quality is crucial for ensuring product safety and reliability. Traditional manual inspection methods are inefficient and subjective, prompting research into automated defect detection technologies. The internet of things (IoT) has revolutionized industrial settings by enabling interconnected devices and sensors, offering new opportunities for real-time and efficient defect detection. This paper proposes a lightweight re-imaging quality assessment method to detect small defects on steel plate surfaces. It uses time-frequency analysis to extract signal frequency characteristics, avoiding "mutual blurring" and enhancing defect features. A lightweight detection model with reduced computational complexity and parameters is developed through depth separation and point-by-point operations, while retaining the ability to capture defect features. Integrating time-frequency spectral features with image domain characteristics allows the network to discern both microscopic and macroscopic information. Empirical validation shows an average accuracy of 95.1% with minimal parameters, and IoT integration improves real-time adaptability, providing a reference for industrial visual inspection tasks.</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

Lightweight re-imaging quality assessment for detecting small defects on steel plate surfaces in the context of the internet of things

  • 期刊论文

  • 发表日期：

    2025年9月17日

查看全文
  • Jingzhao Chen,
  • Bin Liu,
  • Haowen Zuo

摘要

In contemporary industrial production, accurately identifying steel plate surface quality is crucial for ensuring product safety and reliability. Traditional manual inspection methods are inefficient and subjective, prompting research into automated defect detection technologies. The internet of things (IoT) has revolutionized industrial settings by enabling interconnected devices and sensors, offering new opportunities for real-time and efficient defect detection. This paper proposes a lightweight re-imaging quality assessment method to detect small defects on steel plate surfaces. It uses time-frequency analysis to extract signal frequency characteristics, avoiding "mutual blurring" and enhancing defect features. A lightweight detection model with reduced computational complexity and parameters is developed through depth separation and point-by-point operations, while retaining the ability to capture defect features. Integrating time-frequency spectral features with image domain characteristics allows the network to discern both microscopic and macroscopic information. Empirical validation shows an average accuracy of 95.1% with minimal parameters, and IoT integration improves real-time adaptability, providing a reference for industrial visual inspection tasks.