| DC 欄位 |
值 |
語言 |
| DC.contributor | 土木工程學系 | zh_TW |
| DC.creator | 陳青 | zh_TW |
| DC.creator | CING CHEN | en_US |
| dc.date.accessioned | 2009-7-18T07:39:07Z | |
| dc.date.available | 2009-7-18T07:39:07Z | |
| dc.date.issued | 2009 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=963202086 | |
| 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 | 稻作的分布和種植形態對於農政單位是一項重要資訊，使用多時期遙測影像可以用於辨識稻作分布和面積。一般而言，應用長時間系列的多時期衛星影像，基本上會有干擾雜訊，雜訊的來源為感測器內部所產生的系統性雜訊或大氣等外部所影響的隨機性雜訊。因此，時間序列資料可能無法完全反映週期性植物(稻作)時序變化。本研究的主要目的為經驗模態分解法(Empirical Mode Decomposition, EMD)濾除時序雜訊，萃取稻米時序變化的資訊。在時序雜訊濾除後，利用稻米時序變化資訊分類稻米。本研究主要分為四個階段:第一階段，建立常態植生指標(Normalized Difference Vegetation Index, NDVI)時間序列資料。第二階段，以EMD過濾長時間序列資料的雜訊，保留稻米時序變化的資訊。第三階段，在時間序列資料中辨識稻作插秧日期。第四階段，結合相關係數和無母數統計的方法，利用時間序列資料分類稻作。本研究所使用的測試影像為台灣(2005)每8天MODIS /Terra影像。測試結果顯示EMD可以有效濾除NDVI時間序列資料中的雜訊，進而重建稻作時序變化的資訊。
| zh_TW |
| dc.description.abstract | The extent and distribution of individual crop types were important information of government. A number of studies have been carried out using multi-temporal satellite images to identify crop patterns and to classify cropland areas.In general, image noise is any unwanted disturbance in digital image. These noises recorded by the sensor, and are caused by the radiance absorption or emission of atmospheric particles during energy transportation process. As the result, the time-series data do not exactly follow the annual cycles of growth of vegetation, such as paddy rice. In this study, a method using Empirical Mode Decomposition (EMD) to reduce the noise and determine characteristic phenology of paddy rice at a long time scale is proposed. After the EMD filtering, another objective is to mapping the paddy rice with the time series data in Taiwan. The method of this study basically includes four main steps: (i) constructing Normalized Difference Vegetation Index (NDVI) time-series data; (ii) using EMD is proposed to reduce the noise in a year-long MODIS images and retrieve the time variation of paddy rice.; (iii) verifying the planting date of paddy rice from the time series data; (iv) combining of Correlation Coefficient and Corresponding nonparametric test are employed to classify the paddy rice from the time series data. The 8-day MODIS/Terra data in Taiwan (2005) are used to test the proposed method. The results indicate that EMD can effectively reduce noisy components of NDVI time-series profile and the characteristic phenology of rice crop in the region could be constructed from the derived-smoothed NDVI profile.
| en_US |
| DC.subject | 稻米生長期 | zh_TW |
| DC.subject | 時序分析 | zh_TW |
| DC.subject | 經驗模態分解法 | zh_TW |
| DC.subject | MODIS影像 | zh_TW |
| DC.subject | EMD | en_US |
| DC.subject | Time-series analysis | en_US |
| DC.subject | Rice phenology | en_US |
| DC.subject | MODIS | en_US |
| DC.title | MODIS時間序列影像應用於稻米之判釋 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Mapping paddy rice using Multi-Temporal MODIS Images | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |