Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/6165
標題: 灰指甲影像之偵測與分析
Detection and Analysis of Onychomycosis Images
作者: 陳佳慶
Chen, Chia-Ching
關鍵字: active contour model;動態輪廓模型;CIELAB;region growing;entropic thresholding;LUM filters;fractal dimension;CIELAB;像素聚積成長;熵臨界值化;LUM濾波器;碎形維度
出版社: 電機工程學系
摘要: 
摘 要
灰指甲是一種慢性黴菌感染指甲的疾病,研究顯示大約有百分之三到百分之五的人被此疾病所困擾。在亞熱帶的台灣,一年四季都可以發現許多灰指甲的病患。雖然灰指甲對人的健康不會有太大的影響,但是在自尊、工作和走路舒適度方面,它所帶來的一些不適和窘迫可能影響了病患的生活品質。傳統的偵測方式是一維的,醫生用尺測量灰指甲面積的單一向量的長度;這樣的測量方式既不準確,也不客觀。相對於傳統的方法,我們以下討論之方法是二維的,利用影像處理去偵測灰指甲病理影像的面積,而有助於醫生對於用藥方式與劑量的控制。
在本論文中,我們利用動態輪廓模型,去掉大部分的背景,將指甲從病理影像抽取出。之後以CIELAB為基礎,藉由正常指甲與灰指甲之RGB頻帶差異,我們將可以得到一張具有影像增強效果的特徵影像;亦及在此張影像中,灰指甲將顯現較高的灰階值,而正常指甲會顯現較低的灰階值,如此一來有助於後續影像的分割。在分割的方法中,我們使用像素聚積成長、熵臨界值化、LUM濾波器和碎形維度來比較。凸顯之後的影像,變成一個分割的問題。由於帶有紋理的特性,所以我們使用碎形維度來分割灰指甲會有較佳的效能,得到所要的灰指甲面積。經由比較的結果,碎形維度之診斷效能跟醫生一樣,甚至比醫生還好。

Abstract
Onychomycosis is a chronic mycotic infection of nails. Studies estimate that it afflicts three to five percent of the population. In subtropical Taiwan, we can find many sufferers all the year round. Although the adverse impact of onychomycosis on the patient''s overall health status may not be profound, the discomfort and embarrassment associated with the disease affect the patient''s quality of life in terms of self-esteem, social function, and ability to walk comfortably. The conventional method uses 1-D measurement in which method doctors use ruler to calculate the length of clinical image. It's not correct and objective. In opposition to the conventional method, we further discuss the method is 2-D. Using the image processing to recognition the area of clinical onychyomycosis, and the area help doctors how to medicine and control the quantity.
In this paper, we propose active contour model to remove the background and extract the nail plate of the clinical image. After that, we can get an enhanced image that obtain from CIELAB. In this enhanced image, the gray levels of onychyomycosis are brighter than the normal nail. In segmenting methods, we use region growing, entropic thresholding, LUM filters and fractal dimension (FD). Because the area of onychyomycosis is some kind of texture, we employ a technique based on the FD and multi-fractal concept to segment the area of onychyomycosis. After comparing, the FD method diagnoses as good as doctors or better than doctors.
URI: http://hdl.handle.net/11455/6165
Appears in Collections:電機工程學系所

Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.