| DC 欄位 |
值 |
語言 |
| DC.contributor | 土木工程學系 | zh_TW |
| DC.creator | 陳可薰 | zh_TW |
| DC.creator | Ko-Hsun Chen | en_US |
| dc.date.accessioned | 2006-7-13T07:39:07Z | |
| dc.date.available | 2006-7-13T07:39:07Z | |
| dc.date.issued | 2006 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=93322079 | |
| dc.contributor.department | 土木工程學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 傳統上使用變異向量分析法進行變遷偵測，往往有門檻值給定不客觀的問題，雖然有研究學者利用先行圈選訓練區塊的方式來加以改良，但訓練區圈選的好壞，往往會嚴重影響變遷偵測的成果。因此本研究提出一新方法，改良傳統的變異向量分析法。本研究所提出的改良式變異向量分析法，分為以下兩個步驟： (1) 變遷/非變遷像元位置的偵測： 本研究結合變異向量距離與光譜角資訊來進行變遷/非變遷像元的偵測，首先結合光譜角和變異向量距離資訊進行主成分分析，再運用邊緣偵測技術自動化決定第一主成分(PC1)影像的門檻值，即可偵測出變遷與非變遷像元。 (2) 變遷像元類別的判釋: 先計算變異向量之方向角，再運用K-Means演算法對變遷點之方向角進行分類，最後獲得變遷類別的分類影像。本研究利用模擬影像和三組SPOT-5的影像進行測試，在變遷像元偵測的精確度可達95%以上；而在變遷分類的結果，其精確度為87%。實驗測試成果顯示，利用改良式變異向量分析法能有效偵測出影像上變遷像元的位置，並分類出變遷像元的類別。 | zh_TW |
| dc.description.abstract | Change-vector analysis (CVA) traditionally was applied to detect land cover change by thresholding the difference image according to empirical strategies or manual trial-and-error procedure. It made the result of change detection subjective. Recently training areas are used to guide the selection of the change magnitude of the threshold for discriminating Change and No-Change pixels. However, the result would be greatly affected by the appropriately selected training areas. In this paper, a new method is proposed to improve CVA. The modified change-vector analysis consists of two stages. (1) Change/No-Change detection: The method combines the distance of change vector with spectral angle of change vector to detect Change/No Change pixels. In the first place, the distance of change vector and spectral angle were analyzed by Principal Component Analysis. Then, the edge detection technique was used to find out the optimum threshold automatically in the PC1 image. Therefore the Change and No-Change pixels can be detected. (2) Change categories detection: The direction angles of change vector were calculated beforehand. Then, the direction angles of change vector were classified by the K-Means clustering algorithm to determine change category of the change pixels. In this study, the modified change-vector analysis was applied to the detection of one simulation image and three SPOT-5 satellite images. The results indicated that the overall accuracy of Change/No-Change detection was about 95%, and the overall accuracy of change categories detection was 87%. The experimental results indicate that using the modified change-vector analysis is an effective method to detect Change/No-change pixels and change categories. | en_US |
| DC.subject | 變異向量分析法 | zh_TW |
| DC.subject | 變異向量方向角 | zh_TW |
| DC.subject | 主成分分析 | zh_TW |
| DC.subject | 變異向量距離 | zh_TW |
| DC.subject | Change-vector analysis | en_US |
| DC.subject | Principal Component Analysis | en_US |
| DC.subject | the distance of change vector | en_US |
| DC.subject | the direction angles of change vector | en_US |
| DC.title | 改良式變異向量分析法於影像變遷之研究 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Modified Change-vector analysis for Change Images | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |