Study on the Method of Blood Microscopic Multi-field Image Stitching

2019 IEEE 7th International Conference on Bioinformatics and Computational Biology ( ICBCB) Pub Date : 2019-03-01 DOI:10.1109/ICBCB.2019.8854666

Zhangyong Li, Hui Liu, Mengxi Ju, Fuqu Chen, Xinwei Li

{"title":"Study on the Method of Blood Microscopic Multi-field Image Stitching","authors":"Zhangyong Li, Hui Liu, Mengxi Ju, Fuqu Chen, Xinwei Li","doi":"10.1109/ICBCB.2019.8854666","DOIUrl":null,"url":null,"abstract":"In the diagnosis of medical blood diseases, there are contra-dictions between the clear view and the size of view under the microscope. In order to obtain clear blood cell images under a large view, this paper proposes an image stitching method for multi-view blood microscopy images. The method firstly preprocesses the input image sequence, and then uses the SIFT feature and the local LBP feature to extract the feature points of the image sequence, obtains the matching point pairs according to the threshold method, and then uses the improved RANSAC algorithm to calculate the homography matrix between the images. Finally, the weighted average in image fusion is used to realize the seamless stitching of multiview images. The experimental results show that the improved feature detection algorithm has good performance in the rotary image, blurry image and distorted cell image. The improved RANSAC algorithm effectively improves the computational efficiency of the image, and finally achieves multi-view blood display with high efficiency and seamless stitching of micro images.","PeriodicalId":136995,"journal":{"name":"2019 IEEE 7th International Conference on Bioinformatics and Computational Biology ( ICBCB)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE 7th International Conference on Bioinformatics and Computational Biology ( ICBCB)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICBCB.2019.8854666","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

Abstract

In the diagnosis of medical blood diseases, there are contra-dictions between the clear view and the size of view under the microscope. In order to obtain clear blood cell images under a large view, this paper proposes an image stitching method for multi-view blood microscopy images. The method firstly preprocesses the input image sequence, and then uses the SIFT feature and the local LBP feature to extract the feature points of the image sequence, obtains the matching point pairs according to the threshold method, and then uses the improved RANSAC algorithm to calculate the homography matrix between the images. Finally, the weighted average in image fusion is used to realize the seamless stitching of multiview images. The experimental results show that the improved feature detection algorithm has good performance in the rotary image, blurry image and distorted cell image. The improved RANSAC algorithm effectively improves the computational efficiency of the image, and finally achieves multi-view blood display with high efficiency and seamless stitching of micro images.

查看原文本刊更多论文

血液显微多场图像拼接方法的研究

在医学血液病的诊断中，显微镜下视野清晰与视野大小存在矛盾。为了在大视角下获得清晰的血细胞图像，提出了一种多视角血液显微图像的图像拼接方法。该方法首先对输入图像序列进行预处理，然后利用SIFT特征和局部LBP特征提取图像序列的特征点，根据阈值法得到匹配点对，然后利用改进的RANSAC算法计算图像之间的单应性矩阵。最后，利用图像融合中的加权平均实现多视点图像的无缝拼接。实验结果表明，改进的特征检测算法在旋转图像、模糊图像和变形细胞图像中都有良好的检测效果。改进的RANSAC算法有效地提高了图像的计算效率，最终实现了微图像高效无缝拼接的多视角血液显示。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

2019 IEEE 7th International Conference on Bioinformatics and Computational Biology ( ICBCB)

自引率

0.00%

发文量