2010-03-012024-05-17https://scholars.lib.ntu.edu.tw/handle/123456789/684271摘要：本年度計畫期望結合彩色融合衛星影像、航照影像或UAV空拍等遙測影像來源，利用其高時效性、高解析度等特性，透過水稻田判釋系統中已建立的空間資訊判釋、判釋錯誤分布推估及自動人機互動判釋技術，以期能提昇敏感性作物耕作調查之效率及統計正確性，並針對鳳梨與香蕉等敏感性作物，提出粗略與精算等兩種模式之耕作面積判定機制，以因應不同人力需求之單位使用，並落實實際應用。預計針對敏感性作物（香蕉與鳳梨）自動判釋正確性進行提升與不同試做區之特性進行測試。 1.水稻田判釋系統實做擴充及修正，透過水稻田判釋系統中已建立的空間資訊判釋、判釋錯誤分布推估及自動人機互動判釋技術，以提昇敏感性作物耕作調查之效率及統計正確性，並落實實際應用。2.遙測影像自動判釋於不同試做區之適用性評估。3.敏感性作物（鳳梨與香蕉）通報資料勾稽與整合敏感性作物之遙測判釋、ＧＩＳ資料、輔助調查系統及現地調查系統。4.非落於影像判釋敏感性作物之潛勢區域，可做為鄉鎮或縣市單位抽查申報錯誤之重點區域，以提升抽查績效，並有效遏止農戶誤報或錯報。5.在提高判釋的準確度及降低人工依賴程度的原則下，整合出敏感性作物遙測判釋流程與方法，快速且低成本的掌控及預估敏感性作物的生產區位與面積。 <br> Abstract: This project tried to integrate satellite image, aerial photo image or UAV image into rice interpretation system, together with built spatial information, error probability anticipation, and human-machine interactive recognition techniques to upgrade the investigation accuracy and correctness of sensitivity plant for real applications. This year the automatic interpretation system has to be modified upgrade the correctness and different test area for sensitivity include banana and pineapple, the anticipated benefits are: 1. The interpretation system has to make function extension and modification, through the spatial information interpretation, error probability anticipation, and human-machine interactive recognition techniques to upgrade the investigation accuracy and correctness of sensitivity plant for real applications. 2. Remote sensing imagery used for different regions for suitability appraisal. 3. The sensitivity plant announcement records have to be mutual verification with RS interpreted image , GIS data, ancillary investigation system and field checking system. 4. Not within the potential area from RS interpreted image of sensibility plant, the area can be treated as the emphasized area of inspection on announcement records for county and village, in order to promote the achievement in terms of preventing farms&apos; mistake on announcement records. 5.Under the principle of upgrade accuracy of interpretation and diminish the human dependency, integrating interpretation procedures and methods on the sensitivity plant, this research can handle faster and lowers the cost in terms of anticipating the planting position and areas of sensitivity plant.敏感性作物遙測影像判釋sensitive cropremote sensingimage interpretation