李鹏飞,魏一.基于RGB视觉的瓦楞包装褶皱标定收敛研究[J].包装工程,2022,(5):271-276.
LI Peng-fei,WEI Yi.Convergence of Corrugated Board Packaging Wrinkle Calibration Based on RGB Vision[J].Packaging Engineering,2022,(5):271-276. |
| 基于RGB视觉的瓦楞包装褶皱标定收敛研究 |
| Convergence of Corrugated Board Packaging Wrinkle Calibration Based on RGB Vision |
| 投稿时间：2021-05-27 |
| DOI：10.19554/j.cnki.1001-3563.2022.05.037 |
| 中文关键词: 机器视觉 瓦楞纸板包装 RGB视觉 优化分析 |
| 英文关键词: machine vision corrugated board packaging RGB vision optimization analysis |
| 基金项目: |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 对瓦楞纸板包装的褶皱特征精确定位，保证瓦楞纸板包装容器的褶皱特征在生产中被及时检验。方法 机器视觉瓦楞纸板包装检测系统依据图像信息收集、识别算法、RGB视觉、位置标定、收敛性结果对比等部分，确定瓦楞纸板包装褶皱特征标定核心算法要点;在传统频域变换的基础上，进行RGB视觉改进算法分析，有效提高对瓦楞纸板包装的褶皱特征识别率，明确该类包装的检测要点。以实际的瓦楞纸板褶皱为例，论证该类包装褶皱特征的特征提取方法，并且强化研究对象的褶皱特征，最后对识别结果和实物结果进行对照分析。结果 通过设定阙值在90～120得到线性率为25%～80%，结果说明该视觉方式可以实现对机器视觉的瓦楞纸板包装标定，并且各特征的实际回归基准率为86%，验证了RGB视觉算法对褶皱特征标定的精确性。结论 基于该RGB视觉以及优化措施，确定了机器视觉瓦楞纸板包装标定方式，满足了瓦楞纸板包装生产的有效定位要求。 |
| 英文摘要: |
| The work aims to accurately locate the wrinkle features of the existing corrugated board packaging, and ensure that the wrinkle features of corrugated board packaging containers can be timely and effectively inspected in production. According to image information collection, recognition algorithm, RGB vision, location calibration, convergence result comparison, etc., the machine vision corrugated board packaging inspection system determined the key algorithm points of corrugated board packaging wrinkle feature calibration. The RGB visual improvement algorithm was analyzed based on the traditional frequency-domain transformation. The wrinkle feature recognition rate of corrugated cardboard packaging was effectively improved. The testing points of this kind of packaging were clarified. With the actual corrugated board as an example, the feature extraction method of the wrinkle features of the packaging was demonstrated, and the visual features of the wrinkles were strengthened. Finally, the convergence results were compared. The linear rate of passing the threshold of 90~120 was 25%~80%. The results showed that this vision method can realize the corrugated board packaging calibration of machine vision, and the actual regression base rate of each feature was 86%, verifying the accuracy of RGB vision algorithm for wrinkle feature calibration. Based on the RGB vision and optimization measures, the machine vision calibration method of corrugated board packaging is determined, which meets the effective positioning requirements of corrugated board packaging production. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
渝公网安备 50010702501716号