Dissertation
Dissertation > Medicine, health > Clinical > Diagnostics > Laboratory diagnosis > Histology

Research of Automatic Microbe-screening for Liquid-based Cervical Cell Images

Author SunKai
Tutor WangChuanZhong
School Nanchang University of Aeronautics and
Course Control Engineering
Keywords Liquid-Based Cervical Cell Images Automatic Microbe-Screening LevelSet Rough-Set DSP
CLC R446.8
Type Master's thesis
Year 2013
Downloads 23
Quotes 0
Download Dissertation

The problems in the process of automatic microbe-screening for liquid-based cervical cell images:Based on the sample standard,solving four kinds of microbial segmentation and feature-identificating problem,realizing the fast automatic screening.In this paper, do some research in the problems in the process of microbe-screening for liquid-based cervical cell images and use different of image processing algorithms and technology to solve these problems.(1) In the process of automatic microbe-screening for liquid-based cervical cell images, in the paper starting from analysing physical characteristics of microbe,bringing in automatic image testing technology, establishing detection theory frame, putting forward a kind of microbial active-vision solution, making instructional detection.Firstly through the sample to study, establishing sample standard distinguishing cells by microbe-infected from normal cells. Lastlyjudging existence of microbe based on sample relative standard. The experimental results show that this method can assist the doctor’s diagnosis and improve the efficiency.(2) A problem that part of the cervical nucleus by microbe-infected background is complicated and edge is weak in the microscopic images. This paper based on the level set by introducing the liquid-base cervical cells image microbial nuclei segmentation algorithm. The experimental results can be seen that this method can effectively segment nuclei by microbial infected for liquid-base cervical cells image.(3) In the case of overlapped microbial characteristics,microbial intelligent identification problem.This paper introduces a automatic image recognition technology, put forward a kind of automatic identification method based on microbial rough-set model, make instructional recognition.The experimental results show that this method can assist the doctor’s diagnosis and improve the efficiency.(4) According to translantation and optimization problem of auxiliary segmentation algorithm based on DSP.Based on the study of microbe-screening algorithm and DSP image processing technology, constructing embedded platform of automatic microbe-screening system for liquid-based cervical cell images based on DSP. Compared with PC-Matlab environment,this automatic auxiliary segmentation system for liquid-based cervical cell images can rapidly segment the microbial.In this paper, do some applied research in image segmentation、identification、 screening and DSP platform algorithm implementation technology to solve some problems in automatic microbe-screening for liquid-based cervical cell images and achieve the desired purpose.

Related Dissertations
More Dissertations