Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells
碩士 === 朝陽科技大學 === 資訊管理系 === 102 === In Taiwan, cervical cancer is one of the top ten most common cancers among women. It can be detected in the early stage with some screening tool which will reduce the mortality and morbidity by the early treatment. Pap smear is an easy and inexpensive screening to...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2014
|
Online Access: | http://ndltd.ncl.edu.tw/handle/63432812421001041080 |
id |
ndltd-TW-102CYUT0396006 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-102CYUT03960062015-10-13T23:28:40Z http://ndltd.ncl.edu.tw/handle/63432812421001041080 Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells 基於形態學雙構模型的子宮頸抹片細胞分割與分類 Jun-Hao Huang 黃俊豪 碩士 朝陽科技大學 資訊管理系 102 In Taiwan, cervical cancer is one of the top ten most common cancers among women. It can be detected in the early stage with some screening tool which will reduce the mortality and morbidity by the early treatment. Pap smear is an easy and inexpensive screening tool for the cervical cancer screening. However, it is time consuming and has no quantitative criteria. Therefore, the development of an automatic detector for the abnormal cervical cell is thus necessary. This research proposed an effective method, combined with K-means clustering algorithm and a dynamic threshold method based on Dual Reconstruction approach. Then the support vector machine (SVM) classifier is adopted for the discrimination between normal and abnormal cells. In the experimental results, we achieved precision of 87.22% and some examples for the well segmented results. Shao-Kuo Tai 戴紹國 2014 學位論文 ; thesis 56 zh-TW |
collection |
NDLTD |
language |
zh-TW |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 朝陽科技大學 === 資訊管理系 === 102 === In Taiwan, cervical cancer is one of the top ten most common cancers among women. It can be detected in the early stage with some screening tool which will reduce the mortality and morbidity by the early treatment. Pap smear is an easy and inexpensive screening tool for the cervical cancer screening. However, it is time consuming and has no quantitative criteria. Therefore, the development of an automatic detector for the abnormal cervical cell is thus necessary. This research proposed an effective method, combined with K-means clustering algorithm and a dynamic threshold method based on Dual Reconstruction approach. Then the support vector machine (SVM) classifier is adopted for the discrimination between normal and abnormal cells. In the experimental results, we achieved precision of 87.22% and some examples for the well segmented results.
|
author2 |
Shao-Kuo Tai |
author_facet |
Shao-Kuo Tai Jun-Hao Huang 黃俊豪 |
author |
Jun-Hao Huang 黃俊豪 |
spellingShingle |
Jun-Hao Huang 黃俊豪 Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells |
author_sort |
Jun-Hao Huang |
title |
Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells |
title_short |
Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells |
title_full |
Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells |
title_fullStr |
Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells |
title_full_unstemmed |
Segmentation and Classification Based on Morphological Dual Reconstruction in Pap Smear Cells |
title_sort |
segmentation and classification based on morphological dual reconstruction in pap smear cells |
publishDate |
2014 |
url |
http://ndltd.ncl.edu.tw/handle/63432812421001041080 |
work_keys_str_mv |
AT junhaohuang segmentationandclassificationbasedonmorphologicaldualreconstructioninpapsmearcells AT huángjùnháo segmentationandclassificationbasedonmorphologicaldualreconstructioninpapsmearcells AT junhaohuang jīyúxíngtàixuéshuānggòumóxíngdezigōngjǐngmǒpiànxìbāofēngēyǔfēnlèi AT huángjùnháo jīyúxíngtàixuéshuānggòumóxíngdezigōngjǐngmǒpiànxìbāofēngēyǔfēnlèi |
_version_ |
1718085798104924160 |