刘劲松,臧雪颖,陈大勇,等.基于机器视觉的铜管锯齿伤智能识别与软件实现[J].精密成形工程,2025,17(2):130-140.
LIU Jinsong,ZANG Xueying,CHEN Dayong,et al.Intelligent Identification and Software Implementation of Saw-Tooth Injuries on Copper Tubes Based on Machine Vision[J].Journal of Netshape Forming Engineering,2025,17(2):130-140. |
| 基于机器视觉的铜管锯齿伤智能识别与软件实现 |
| Intelligent Identification and Software Implementation of Saw-Tooth Injuries on Copper Tubes Based on Machine Vision |
| 投稿时间：2024-09-12 |
| DOI：10.3969/j.issn.1674-6457.2025.02.015 |
| 中文关键词: 精密铜管 机器视觉 锯齿伤 缺陷识别 缺陷占比 |
| 英文关键词: precision copper tubes machine vision saw-tooth defects defect identification defect proportion |
| 基金项目:辽宁省教育厅基本科研项目(LJKMZ20220591);常州市领军型创新人才引进培育项目(CQ20220057) |
|
| 摘要点击次数: 264 |
| 全文下载次数: 23 |
| 中文摘要: |
| 目的 锯齿伤作为铜管材首要的缺陷，会造成管材后续二次加工开裂和服役破裂等严重问题，对缺陷进行高效、定量化的统计及分析是产品质量管控的关键性难点，也是铜管加工行业的空白。针对这一问题，基于机器视觉方法，设计并开发了铜管材锯齿伤缺陷智能识别算法和系统。方法 首先，基于自适应阈值算法，设计采用一系列的降噪算法，实现锯齿伤管材横截面轮廓识别。其次，定义了锯齿伤缺陷占比的计算规则，通过算法优化，实现了开口和闭口2种锯齿伤的成功识别和缺陷占比的计算。最后，借助Python语言构建了图形用户界面(GUI)系统，完成了锯齿伤缺陷统计的软件化。结果 通过对智能识别算法和实验获得的锯齿伤缺陷占比进行对比，发现相对误差保持在1%~2%，证明缺陷智能识别算法具有可行性和高精度。结论 该系统可以有效地实现铜管锯齿伤缺陷识别和占比的定量、高效统计与分析。 |
| 英文摘要: |
| Serration defects, as the primary flaw in copper tubing, can lead to severe issues such as cracking during secondary processing and in-service failures. Efficient and quantitative statistical analysis of these defects is a critical challenge in quality control and a gap in the copper tube processing industry. The work aims to design and develop an intelligent defect recognition algorithm and system for copper tube serration defects based on machine vision methods to address this issue. First, using an adaptive threshold algorithm, a series of denoising algorithms were implemented to achieve the identification of the cross-sectional profile of tubes with serration defects. Subsequently, calculation rules for the proportion of serration defects were defined. Through algorithm optimization, both open and closed serration defects were successfully identified and quantified. Finally, a Graphical User Interface (GUI) system was constructed with the help of the Python programming language, achieving the software implementation of serration defect statistics. By comparing the serration defect proportions obtained from the intelligent recognition method with experimental data, the relative error was found to be 1% to 2%, proving the feasibility and accuracy of the defect recognition algorithm. This system can effectively realize the quantitative and efficient statistical analysis of serration defects in copper tubes. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
渝公网安备 50010702501719号