张绪红,陈滢,张先勇,刘南渤.点特征提取在锂电池丝印图像配准的应用[J].包装工程,2021,42(5):290-298.
ZHANG Xu-hong,CHEN Ying,ZHANG Xian-yong,LIU Nan-bo.Application of Point Feature Extraction in Battery Screen Printing Image Registration[J].Packaging Engineering,2021,42(5):290-298. |
| 点特征提取在锂电池丝印图像配准的应用 |
| Application of Point Feature Extraction in Battery Screen Printing Image Registration |
| 投稿时间：2020-08-01 |
| DOI：10.19554/j.cnki.1001-3563.2021.05.038 |
| 中文关键词: 点特征 图像配准 特征提取 性能评价 |
| 英文关键词: point feature image registration feature extraction performance evaluation |
| 基金项目:广东省自然科学基金(2018A0303130111);广州市科技计划(201902020003) |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 为了提高锂电池丝印图像配准精度，从而解决产品质量检测中的漏检和误报问题，研究点特征提取算法在锂电池丝印图像配准中的应用。方法 对基于点特征的锂电池丝印图像配准进行综述，首先概述点特征提取算法的发展历程，然后着重围绕Harris，SIFT，SURF，ORB和AKAZE等5种经典的点特征提取算法进行分析，并介绍近几年的提升算法，最后对锂电池丝印图像进行配准测试，利用几种测评技术对实验效果进行分析，总结不同点特征提取算法在锂电池丝印图像配准中的优缺点和适用性。结果 实验结果表明，AKAZE算法提取的特征点具有较高的重复率和匹配准确率，经过配准后的定位误差也都控制在1个像素以内，但是该算法的尺度不变性较差。结论 相较于前4种算法，AKAZE算法具有较高的可靠性和稳定性，能够满足锂电池丝印图像配准的实时性和高效性需求。 |
| 英文摘要: |
| The work aims to study the application of point feature extraction algorithm in lithium battery screen printing image registration, in order to improve the accuracy of lithium battery screen printing image registration and solve the problems of missing detection and false alarm in the quality inspection. The lithium battery screen printing image registration based on point feature was summarized. Firstly, the development of point feature extraction algorithm was concluded. Then, five classical point feature extraction algorithms including Harris, SIFT, SURF, ORB and AKAZE were analyzed and their improved algorithms in recent years were introduced. Finally, the battery screen images were tested for registration and experiment effect was analyzed by several evaluation technologies to summarize the advantages, disadvantages and applicability of different point feature extraction algorithms in lithium battery screen printing image registration. The point feature extracted by AKAZE had higher repetition rate and matching accuracy rate and the positioning error after registration was controlled within one pixel, but the scale invariance of AKAZE algorithm was poor. Compared with the other 4 algorithms, AKAZE algorithm has high reliability and stability, which can meet the real-time and efficient requirements of lithium battery screen printing image registration. |
|
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