Using Watershed Algorithm in Independent Component Image Analysis

勤益科大機構典藏 > 電資學院 > 資訊工程系(所) > 【資訊工程系所】期刊論文 > Item 987654321/6240

請使用永久網址來引用或連結此文件: http://ir.lib.ncut.edu.tw/handle/987654321/6240

題名:	Using Watershed Algorithm in Independent Component Image Analysis
作者:	游正義
貢獻者:	資訊工程系
日期:	2013-06
上傳時間:	2017-10-02 09:26:11 (UTC+8)
摘要:	As the image quality and computer computing power develop, the un- modifiable physical storage mode of images has turned into modifiable digital storage. However, although the present digital preservation technology of images develops increasingly, there are still many unpredictable variables influencing the preservation quality of a digital image. The quality usually depends on the usage mode of the user or the variables resulted from natural environment. Therefore, if the brightness distribution of an image after image analysis is too extreme, the formed subject matter will be difficult to be recognized. This study combined Independent Component Analysis (ICA) with watershed algorithm, and carried out principal component analysis for the R, G and B color planes of the imported color digital images, in order to achieve better effect on the detection and tracking of primary target than the previously proposed ICA. Experiment results indicate that our approach has great improvement on the images information interpreted and recognized. © 2013 ISSN 2185-2766. (12 refs)
關聯:	ICIC Express Letters, Part B: Applications
顯示於類別:	[資訊工程系(所)] 【資訊工程系所】期刊論文

文件中的檔案:

沒有與此文件相關的檔案.

檢視Licence

在NCUTIR中所有的資料項目都受到原著作權保護.