Skip navigation
  • 中文
  • English

DSpace CRIS

  • DSpace logo
  • 首頁
  • 研究成果檢索
  • 研究人員
  • 單位
  • 計畫
  • 分類瀏覽
    • 研究成果檢索
    • 研究人員
    • 單位
    • 計畫
  • 機構典藏
  • SDGs
  • 登入
  • 中文
  • English
  1. National Taiwan Ocean University Research Hub

Counting the quality and quantity of agricultural land, and establishing an agricultural space information collaboration platform

瀏覽統計 Email 通知 RSS Feed

  • 簡歷

基本資料

Project title
Counting the quality and quantity of agricultural land, and establishing an agricultural space information collaboration platform
Code/計畫編號
109農科-16.1.1-子-i1
Translated Name/計畫中文名
盤點農地的質與量,規劃農業生產資源與調整利用
 
Funding Organization/主管機關
Council of Agriculture,Executive Yuan
 
Co-Investigator(s)/共同執行人
劉頂立(計畫主持人)
梁德容
王尉任
鄭絜文
陳偉文
郭鴻裕
郭志良
吳佳穎
 
Department/Unit
Council of Agriculture,Executive Yuan
Website
https://www.grb.gov.tw/search/planDetail?id=13670749
Year
2020
 
Start date/計畫起
01-01-2020
Expected Completion/計畫迄
31-12-2020
 
Co-Investigator(s)
Chin-Chun Chang
Bugetid/研究經費
96156千元
 
ResearchField/研究領域
農業經濟
 

Description

Abstract
一、現今農業的發展和政策的推動,往往需要依靠大量的數據和機器學習的輔助,進行資訊的取得和分析,有鑑於此,農業委員會委託農業科技研究院農業政策研究中心建構農業施政資料庫,彙整農委會及所屬機關農業業務資料庫。本計畫定期進行資料庫抄錄、備份與整合,提供給各業務單位使用,作為政策研析之用。此外,本計畫深感人工智慧的蓬勃發展,有助於農業的應用,提出以深度學習輔助農作物航空影像判釋,透過多種CNN架構模型訓練,找出最適合的判釋模型,協助判釋人員進行航照判釋。此外,本研究團隊中央大學建立一個深度學習平台,整合影像分割方法、深度學習模式及重新學習程序。使用者可以上傳整張遙測影像,系統判釋後產生農作物區域的邊界資訊及分類結果。判釋結果往往還需再由專家去判斷結果是否正確,使深度學習模型辨識率提升。若判釋錯誤則需修改影像中農作物的邊界區域及分類,讓模式重新學習。如果單以人工的方式建立判釋錯誤的樣本,會花費許多的時間。本系統能輔助專家對遙測影像進行快速標記,然後讓深度學習模式重新學習。二、中國大陸自107年8月3日起持續發生非洲豬瘟疫情,迄108年10月中旬亞洲地區已有10個國家遭受入侵,疫情擴散迅速,由於豬隻感染非洲豬瘟後致死率幾乎100%,造成重大經濟損失,影響民生食肉需求及相關產業生計,遂建立早期預警機制、整備國內飼養管理及屠宰衛生運輸管理等作業以因應各項緊急狀況實為必要之措施。 為防範非洲豬瘟及管制疫情,行政院農業委員會暨行政院農業委員會動植物防疫檢疫局,修正發布之「動物運送管理辦法」及「屠宰作業準則」,強制載運活豬、豬屠體、內臟及其分切物之運輸車輛加裝即時追蹤系統(GPS)車機,即時透過行動網路傳送至車機所屬之車輛軌跡管理系統追查來源牧場與肉品流向。本案係開發共通即時傳輸功能,提供不同廠商之車輛軌跡管理系統即時轉抛即時運輸車輛串流資訊,再運用既有畜牧場地理資訊,配合農委會防檢疫需要,開發即時運輸車輛追蹤、查核與告警等相關功能,以及歷史軌跡處理支援以場追車,以車查場等功能,支援非洲豬瘟防疫指揮管制任務。三、行政院農業委員會暨所屬機關及全國各農田水利會為有效管理全國農漁牧地、山坡地及森林土地,運用地理資訊系統軟體及技術,輔助多項業務執行。ESRI ArcGIS系列軟體因其優良運算效能與分析功能,以及長期的推廣成果,目前為農業委員會推動GIS相關業務之主要應用軟體工具。本計畫預計執行的五大工作項,如下所示: (1)提供既有ESRI ArcGIS地理資訊系統應用軟體之原廠使用授權:本專案期間將提供農委會主管機關(構)既有ESRI ArcGIS地理資訊系統各品項應用軟體相同數量原廠使用授權,不限既有軟體現行版次,並依機關(構)需求升級至該軟體最新版次或提供現行版次之軟體維護服務,有效降低軟體升級維護成本。(2)專案期間無償提供軟體試用授權:依農委會與所屬機關需求,提供ESRI ArcGIS地理資訊系統各品項應用軟體試用授權,有效節省軟體購置成本。(3)製作數位課程:規劃系統性的數位課程,課程內容將包含ArcGIS Pro基礎/進階課程、ArcGIS Enterprise基礎/進階課程、影像處理軟體、相關業務專班等課程,提供學員數位學習,讓課程內容貼合學員業務上之使用或達成特殊任務。(4)提供技術與諮詢服務: 提供ArcGIS軟體升級、安裝等技術服務、軟體應用諮詢服務,有效解決業務單位軟體安裝、操作問題。(5)維運地理資源雲(https://gisportal.coa.gov.tw/portal)平台。 1.In the present, the development and promotion of agricultural policies often need to rely on large amounts of data and machine learning in data analysis. For this reason, the Council of Agriculture (COA) commissioned the Agriculture Policy Research Center of Agricultural Technology Research Institute (ATRI) to construct the agricultural policy database integrate agricultural databases from COA and subsidiary agencies. This project copy, backup, and integrate agricultural databases periodically for policy research analysis purposes. Besides, this project aware of the importance of the development of artificial intelligence. We propose depth learning techniques to assist experts in crop interpretation. Besides, this collaborative project to establish a depth learning platform that integrates the image segmentation method, a deep learning model, and re-learning procedures by National Central University. The user can upload the entire remote-sensing image. After the system's interpreting, it will produce the boundary information and classification results of the crop area. Interpretation results often require experts to judge whether the results are correct, which can improve the recognition rate of the deep learning model. If the results are incorrect, the experts should modify the boundary and class information of the crop area, and the system will re-train the CNN model again. It would take a lot of time to create a wrong interpretation sample manually. The developed system can help experts re-label the remote-sensing image rapidly and then activate deep learning model to re-train.2.The African swine fever epidemic has continued to occur in mainland China since August 3, 107. As of mid October 108, ten countries in Asia have been invaded, and the epidemic has spread rapidly. The death rate of pigs infected with African swine fever is almost 100%. As a result, it has caused significant economic losses and affected livelihood and meat needs and related industries’livelihoods. Therefore, it is necessary to establish an early warning mechanism, set up domestic feeding management and slaughtering sanitary transportation management to respond to various emergency situations. To prevent swine fever in Africa and control the epidemic situation, the Agriculture Commission of the Executive Yuan and the Animal and Plant Quarantine and Quarantine Bureau of the Executive Yuan Agricultural Committee have amended the Administrative Measures for the Transport of Animals and the Slaughtering Guidelines to compulsorily carry live pigs, pig carcasses, internal organs and The cut vehicle is equipped with a real time tracking system (GPS) vehicle, which is transmitted to the vehicle trajectory management system of the vehicle through the mobile network in real time to track the source pasture and meat flow. This case is the development of a common real time transmission function, which provides the vehicle trajectory management systems of different manufacturers for real-time transfer of real time transport vehicle stream information, and then uses the existing geographic information of livestock farms to meet the needs of the Agricultural Commission for quarantine, and develop real time transport vehicle tracking Relevant functions such as inspection, alarm, and historical trajectory processing support functions such as chasing cars on the field and checking factories to support African swine fever epidemic prevention and control tasks.3.For efficient land management of the agricultural, forestry, fishery, and husbandry, The Council of Agriculture (COA) Associates of Executive Yuan, including regional and local Farm Irrigation Associations make use of GIS system software tools to support and execute their multi-task operations. By excellent computing capacity, analysis functionality, and long-term promotion of ESRI GIS software to generate the best performance, ESRI ArcGIS becomes the main GIS software tools to execute its related tasks in the COA Executive Yuan. As planned, there are five main items to be executed as follows: (1))Maintaining existing application software licenses of ESRI ArcGIS: During the project period, Users that have existing application software licenses of ESRI ArcGIS can continue to use original version or upgrade the latest version on demand. This can reduce maintenance costs.(2))Provided trial licenses of ESRI ArcGIS without extra payment during project delivery period: Users can apply for trial licenses of ESRI ArcGIS on demand. This can save purchasing costs.(3))Offered ESRI ArcGIS software training course: We provide ArcGIS Pro basic and Advanced course, ArcGIS Enterprise basic and Advanced courses, Drone2Map courses etc., let users complete their daily tasks more efficiently. And providing ArcGIS professional courses to any unit  of COA, let the course content fit the users.(4))Provided technical and consulting service: Solve the problem effectively about software installation, upgrade or using.(5))Maintaining and Operating the ArcGIS Enterprise cloud platform (https://gisportal.coa.gov.tw/portal).
 
Keyword(s)
全球定位系統
運輸車輛
非洲豬瘟
GPS
transport vehicle
African swine fever
 
瀏覽
  • 機構典藏
  • 研究成果檢索
  • 研究人員
  • 單位
  • 計畫
DSpace-CRIS Software Copyright © 2002-  Duraspace   4science - Extension maintained and optimized by NTU Library Logo 4SCIENCE 回饋