摘要
目的开发不同的计算方法,对图像拼接的质量进行评价,为逐步建立医疗器械检验检测中图像质量评价的规范提供依据。方法以显微镜成像为例,在原始图像内部、粗糙拼接图像边界和改善后的拼接图像边界上分布计算了拼接图像边界的二范数误差和相关系数,通过散点图观察计算结果的分布。结果图像边界的二范数误差和相关系数的分布可用于区分粗糙拼接和平滑拼接。结论本文研究了图像拼接面临的问题和量化评价方法,提出了用分类方法评价拼接质量的可行性,为探索客观的评价指标提供参考。
Objective To establish standard method for providing information to assess image stitching in medical device testing based on developing different calculation methods to evaluate the quality of image stitching.Methods Taking microscopy as an example,this paper calculated2-norm and correlation coefficients on the borders of original image interior border,roughly stitched image and improved stitched image.The results were shown in scatter plots.Results The distribution of2-norm and correlation coefficients on the image border could be used to differentiate rough stitching and smooth stitching.Conclusion This paper studied the problems and quantified evaluation method,and proposed the feasibility to use classification method for evaluation.It provides reference for further exploration of quantified evaluation parameters.
作者
王浩
孟祥峰
刘艳珍
任海萍
WANG Hao;MENG Xiang-feng;LIU Yan-zhen;REN Hai-ping(Division of Active Medical Device and Medical Optics, National Institutes for Food and Drug Control, Beijing 100050, China)
出处
《中国医疗设备》
2017年第8期20-23,32,共5页
China Medical Devices
基金
国家科技支撑计划项目(2015BAI43H00)
中国食品药品检定研究院中青年基金项目(2015C02)
关键词
医疗器械
质量控制
医学成像
图像处理
medical device
quality control
medical imaging
image stitching
