查詢結果分析
相關文獻
- 協合ERS-1 SAR 及 Landsat TM 影像於埔里地區地覆分類之初探
- 三種尺度遙測影像於林地被覆分類系統之建立
- 雙都普勒雷達熱動力反求法之改進--應用於颮線與環境交互作用之動力結構分析
- VHF陣列雷達大氣回波來向角之定位偏差--雷達波束場型之不均勻加權效應
- 雷達衛星輔助工程環境監測--開放資料的潛力與機會
- Detection for High Resolution Radars
- 應用合成口徑雷達干涉量測技術偵測大地微量變化
- 地理資訊系統、遙感探測及衛星定位系統在坡地管理之應用
- CPT與GPR在大地工程之應用及展望
- 梅雨鋒附近中尺度渦旋之雷達回波與風場特徵: 1997年6月2日個案分析
頁籤選單縮合
| 題 名 | 協合ERS-1 SAR 及 Landsat TM 影像於埔里地區地覆分類之初探=Preliminary Study on the Synergism of ERS-1 SAR and Landsat TM Images for Land Cover Classification over Pu-Li Area |
|---|---|
| 作 者 | 黃凱易; 蘇婉霞; 李旻旻; | 書刊名 | 國立中興大學實驗林研究彙刊 |
| 卷 期 | 19:1 1997.03[民86.03] |
| 頁 次 | 頁145-162 |
| 分類號 | 440.98 |
| 關鍵詞 | 遙感探測; 雷達; 多譜影像分類; Remote sensing; Radar; Multispectral image classification; |
| 語 文 | 中文(Chinese) |
| 中文摘要 | 本研究以埔里地區之ERS-1 SAR與Landsat TM組合影像完成多頻譜分類 ,旨在了解ERS-1 SAR影像資料可否有效改善分類準確度。本研究對ERS-1 SAR的C波段影像初步判釋, 建立地覆雷達反射特性表, 然後以最大相似法分別對 Landsat TM 影像及組合影像分類。 SAR C波段影像判釋結果顯示僅以灰度值(色調) 無法有效分辨地覆。 Landsat TM 最佳分類波段組合為1、3、4、7,分類準確度約83 %,而SAR C波段與前者之組合影像的分類準確度與前者相近,但未能有效改善分類之 準確度。此乃歸因於ERS-1之高俯角設計導致嚴重地形效應;單一頻率之C波段對試區 地覆分辨能力不足,且此頻率可能不是試區內所有地覆最佳選擇。因此,若能使用多頻率、 多俯角、多偏振雷達資料,極可能改善分類準確度,惟需慎選波段組合。故後續研究可在探 討如何確認一最佳雷達波段組合。 |
| 英文摘要 | The objective of this study was to investigate if the synergism of ERS -1 SAR and Landsat TM images could effectively improve classification accuracy. Preliminary interpretation of ERS-1 SAR image was conducted in order to establish a look-up table for typical surface features and their signatures on the SAR image. Also multispectral classifications using the maximum likelihood classifier were performed both on the composite image and only on the Landsat TM image. The discrimination ability of ERS-1 SAR for land cover classification was shown to be lower than that of each TM band (excluding TM band 6). The band combination of TM1,3,4 and 7 was optimal with a classification accuracy of 83%; however, the classification accuracy of the composite image was somewhat lower than that value. Therefore, the classification accuracy could not be effectively improved with the addition of SAR image. The discrimination ability of ERS-1 SAR for land cover classification was lower primarily due to the topographic effects introduced by high depression angle of ERS-1 SAR over mountainous terrain. The second reason is possible that the satellite operates on a single frequency and this frequency may be not optimal for land cover classification over Pu-Li area. There is a high possibility to improve classification accuracy if multi-frequency, multi-polarization, multi-depression angle SAR data are used; however, an appropriate selection of band combination is required. The follow-up studies will be to investigate how to determine an optimal SAR band combination. |
本系統中英文摘要資訊取自各篇刊載內容。