Scholarlink

The detection of surface defects in chemical tanks is both challenging and costly for industrial quality control. Manual inspections are inefficient and prone to errors, failing to meet modern industry’s demands for precision and speed. This study presents an automated inspection system utilizing drones and machine vision. Equipped with high-resolution cameras, light regulators, and a stable drone platform, the system ensures comprehensive tank surface coverage. By integrating the ResNet-50 deep learning algorithm, it achieves precise detection of defects like cracks and corrosion. The proposed solution enhances efficiency, reduces manpower reliance, and holds significant industrial potential. Future work includes model optimization, multi-sensor integration, and improved lighting controls.

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

Research on a Machine Vision-Based Chemical Tank Surface Defect Detection System Using Drone Inspection

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Jiawei Xiang,
  • Yuwei Xiang,
  • Yuming Zhou

摘要

The detection of surface defects in chemical tanks is both challenging and costly for industrial quality control. Manual inspections are inefficient and prone to errors, failing to meet modern industry’s demands for precision and speed. This study presents an automated inspection system utilizing drones and machine vision. Equipped with high-resolution cameras, light regulators, and a stable drone platform, the system ensures comprehensive tank surface coverage. By integrating the ResNet-50 deep learning algorithm, it achieves precise detection of defects like cracks and corrosion. The proposed solution enhances efficiency, reduces manpower reliance, and holds significant industrial potential. Future work includes model optimization, multi-sensor integration, and improved lighting controls.