基于图像拼接的医疗器械高精度检测技术研究

VIP免费

3.0 陈辉 2024-11-19 4 4 2.53MB 56 页 15积分

侵权投诉

摘要

随着医疗卫生科学的不断进步，各种医疗器械在对病人的辅助治疗与辅助诊

断方面发挥着极其重要的作用。对器械进行合格度检测也成为降低对病人伤害的

必要步骤。特别是随着微创手术的日渐增多，大量的微型手术器械通过导管植入

体内进行手术，其尺寸的微小差别可能给病人带来致命伤害。在这种情况下，对

医疗器械特别是微型医疗器械检测显得极为重要。传统的检测方式为接触式测量，

不但效率低而且精度易受人为因素干扰。

机器视觉技术以准确、快速、非接触式等优点广泛应用到现代工业领域，同

时也给医疗器械的非接触式检测提供了可行性方案。利用机器视觉可以检测出药

品罐装时人眼无法观察到的微小杂质、可以检测出一次性注射器针尖部位可能产

生的毛刺以及反表现象(针尖部位误装入针座)等。

但是从精密医疗器械检测角度考虑，由于受限于图像传感器分辨率，常规的

机器视觉检测手段无法实现对精密医疗器械的微米级检测。针对该现象，本课题

提出了基于图像拼接的精密医疗器械高精度检测技术，通过对精密医疗器械图像

进行拼接提高图像分辨率，进而达到提高检测精度的目的。课题设计了基于图像

拼接的医疗器械高精度检测系统，从硬件以及软件方面对影响拼接与检测精度的

因素进行了研究，具体研究方面如下：

（1）在归纳总结有关光学、机器视觉以及图像拼接应用的相关领域与国内

外相关成果后，结合医疗器械检测精度要求高的特点设计出一套包含光源系统、

夹具系统、图像采集与处理系统的机器视觉系统。

（2）分别采用双目图像传感器方式以及步进电机控制方式进行图像采集，

分析两种方式的优缺点，从硬件角度分析两种方式对高精度图像拼接造成的影响

以及改进方案。

（3）分析常规图像拼接算法，对不同算法进行测试，分析算法缺陷并改进，

使其适用于医疗器械图像的高精度拼接。

（4）按检测要求对拼接后的图像进行高精度检测，判断器械合格率。

通过实验证明，通过对系统软硬件进行改进，图像拼接精度有极大提高，对

拼接后的图像进行长度、宽度、角度等参数检测，结果与单目图像传感器检测结

果比较，精度极大提高。基于图像拼接的高精度医疗器械检测技术提供了一种提

高图像分辨率与检测精度的新方法，它克服了对图像传感器的限制，降低了系统

成本，同时该技术适用于对超过视场范围的大型器械进行拼接检测，是一种适合

推广的新技术。

关键字：机器视觉图像拼接双目图像传感器形态学-Harris 算法

高精度检测

ABSTRACT

With the continuous progress of medical science, all kinds of medical equipments

play an extremely important role in the aided diagnosis and aided treatment of patients.

It has become a necessary step for inspection of equipments’ eligible to reduce patient’s

sufferings. Especially with the increasing number of minimally invasive surgery, a large

number of micro-surgical instruments implant into body through the catheter during

surgery, the small difference of size may give the patient a fatal injury. In this case, the

medical device, in particular micro medical device testing is extremely important. The

traditional detection methods have to contact the equipments which were tested, not

only inefficient and the precision is vulnerable to man-made interference.

Machine vision technology is widely used in modern industrial areas; it’s accurate,

rapid and non-contact. While providing a non-contact detection project for medical

devices testing. Using machine vision we can detect impurities that human eye cannot

observe in drug can; can detect the glitches and the phenomenon of anti-form (needle tip

position into the seat) of disposable syringe needle and so on.

But in high precision testing of medical device, due to the limited resolution of the

image sensor, conventional machine vision detection methods cannot achieve micron

precision testing of medical devices. For this, the project of high-precision detection of

precision medical equipments based on the image mosaic is proposed, through mosaic

image of precision medical equipment we can increase image resolution, so as to

achieve the purpose of improving detection accuracy . System of high-precision medical

device testing based on Image mosaic was designed, do much research about influence

factor of measurement accuracy from hardware and software aspects, specific studies

are as follows:

(1) Summarized the domestic and foreign achievements in related fields of optics,

machine vision and image stitching applications, and then combine the

characteristics of high precision medical device testing to design the machine

vision system which includes lighting system, clamping systems, image

acquisition and processing system.

(2) In this project, using two methods for image acquisition: binocular image

sensors and stepper motor control for image acquisition, analysis the

advantages and disadvantages of two methods, from a hardware point of view

to discuss the influence of high-precision image mosaic and improving

programs of these methods.

(3) Analysis of conventional image mosaic algorithm, testing of different

algorithms, and analysis defect of algorithm and improved, making it suitable

for high-precision medical device image mosaic.

(4) According to testing requirements to detect the equipment after high-precision

image stitching to determine eligible rates.

The experimental results show that image mosaic accuracy has been greatly

improved after improvements of system hardware and software. The accuracy of length,

width, angle and other parameters after the image mosaic are greatly enhanced

compared to the monocular image sensor. The detection technology which based on

Image mosaic provide a new method to improve image resolution and measurement

accuracy of high-precision medical device, which overcomes the limitations of the

image sensor, reducing the system cost, while suitable for large equipments detection, is

a suitable new technology for promotion.

Key Word ：Machine Vision, Feature Recognition, Shape Matching,

Curve Fitting, Edge Detection

中文摘要

ABSTRACT

第一章绪论 ...................................................................................................................1

§1.1 机器视觉概述 ...................................................................................................1

§1.1.1 机器视觉技术起源与发展 .....................................................................1

§1.1.2 机器视觉系统构成及应用领域 .............................................................1

§1.2 图像拼接技术概况 ...........................................................................................3

§1.2.1 国内外研究现状 .....................................................................................3

§1.2.2 图像拼接分类 .........................................................................................4

§1.2.3 图像拼接流程 .........................................................................................5

§1.3 课题来源及意义 ...............................................................................................6

§1.4 研究思路和主要内容 .......................................................................................7

第二章高精度图像拼接系统理论 ...............................................................................10

§2.1 医疗器械图像高精度拼接检测流程 .............................................................10

§2.2 图像预处理技术 .............................................................................................11

§2.3 特征点提取与匹配 .........................................................................................12

§2.3.1 基于 Harris 算子的特征点提取 ...........................................................12

§2.3.2 特征点匹配 ...........................................................................................14

§2.4 图像融合 .........................................................................................................15

第三章图像拼接系统介绍 ...........................................................................................17

§3.1 图像拼接视觉系统构成 .................................................................................17

§3.2 图像采集设备 .................................................................................................18

§3.3 光照系统 .........................................................................................................20

§3.4 图像拼接与图像后处理 .................................................................................22

§3.5 辅助装置 .........................................................................................................23

§3.6 系统整体构建 .................................................................................................24

第四章图像拼接实验与精度分析 ...............................................................................28

§4.1 系统标定 .........................................................................................................28

§4.2 步进电机控制微小器械图像拼接 .................................................................29

§4.2.1 微小器械图像拼接 ...............................................................................29

§4.2.2 器械边缘亚像素提取 ...........................................................................31

§4.2.3 器械宽度检测 .......................................................................................33

§4.3 步进电机控制大尺寸器械图像拼接 ..............................................................34

§4.4 基于双目图像传感器的高精度拼接 .............................................................35

§4.4.1 双目传感器图像拼接 ............................................................................36

§4.4.2 圆形拼接图像半径检测 ........................................................................37

§4.5 双目图像传感器方式与步进电机控制方式拼接精度比较 .........................41

§4.6 重叠范围对拼接精度的影响 .........................................................................42

§4.7 软件实现 .........................................................................................................45

第五章总结 .................................................................................................................46

参考文献 .........................................................................................................................48

在读期间公开发表的论文和承担科研项目及取得成果 .............................................51

致谢 ...............................................................................................................................52

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

15 积分 4人已下载

立即下载 VIP免费下载

摘要：

摘要随着医疗卫生科学的不断进步，各种医疗器械在对病人的辅助治疗与辅助诊断方面发挥着极其重要的作用。对器械进行合格度检测也成为降低对病人伤害的必要步骤。特别是随着微创手术的日渐增多，大量的微型手术器械通过导管植入体内进行手术，其尺寸的微小差别可能给病人带来致命伤害。在这种情况下，对医疗器械特别是微型医疗器械检测显得极为重要。传统的检测方式为接触式测量，不但效率低而且精度易受人为因素干扰。机器视觉技术以准确、快速、非接触式等优点广泛应用到现代工业领域，同时也给医疗器械的非接触式检测提供了可行性方案。利用机器视觉可以检测出药品罐装时人眼无法观察到的微小杂质、可以检测出一次性注射器针尖部位可能产生的毛...

展开>> 收起<<

基于图像拼接的医疗器械高精度检测技术研究.pdf

共56页,预览6页

还剩页未读，继续阅读

基于图像拼接的医疗器械高精度检测技术研究

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

推荐作者

热门标签

举报选择: