| 基于彩色纹理分析和神经元网络的结肠镜图像异常检测 |
| |
| 引用本文： | 夏顺仁,陈希超,朱丹琦.基于彩色纹理分析和神经元网络的结肠镜图像异常检测[J].计算机工程,2004,30(9):174-176. |
| |
| 作者姓名： | 夏顺仁 陈希超 朱丹琦 |
| |
| 作者单位： | 浙江大学生物医学工程教育部重点实验室,杭州,310027 |
| |
| 基金项目： | 国家自然科学基金资助项目 (60272029) |
| |
| 摘 要： | 提出了一种新的基于彩色纹理分析和多层前馈型神经元网络的方法，对结肠镜图像进行异常检测。通过对两种常用的灰度图像纹理分析方法(同现矩阵方法和离散小波变换方法)进行扩展，提出相应的彩色纹理分析方法。彩色纹理分析提取出的特征参数构成作为识别器的多层前馈型神经元网络的输入。实验结果表明该方法能有效检测结肠镜图像中的各种异常。
|
| 关 键 词： | 彩色纹理分析 神经元网络 结肠镜图像 异常检测 |
| 文章编号： | 1000-3428(2004)09-0174-03 |
Abnormality Detection in Colonoscopic Images Based on Colour Texture Analysis and Neural Network |
| |
| XIA Shunren,CHEN Xichao,ZHU Danqi.Abnormality Detection in Colonoscopic Images Based on Colour Texture Analysis and Neural Network[J].Computer Engineering,2004,30(9):174-176. |
| |
| Authors: | XIA Shunren CHEN Xichao ZHU Danqi |
| |
| Abstract: | This paper presents a new scheme for abnormality detection in colonoscopic images based on colour texture analysis and neural network. Two colour texture analysis methods are proposed by extending two grey scale methods respectively, i.e. cooccurrence matrix and discrete wavelet transform. Experimental results indicate that this system is able to detect various abnormality in colonoscopic images. |
| |
| Keywords: | Colour texture analysis Neural network Colonoscopic images Abnormality detection |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
