Fisheries Data Analysis, Big Data Visulization, and Applications Based on the Next Generation Oceanic Information System.(III)

瀏覽統計 Email 通知 RSS Feed

簡歷

基本資料

Project title

Code/計畫編號

MOST107-2221-E019-037-MY3

Translated Name/計畫中文名

新世代海洋漁業資訊整合系統之應用, 漁業資料分析, 與巨量資料視覺化(III)

Project Coordinator/計畫主持人

William Wei-Yuan Hsu

Funding Organization/主管機關

National Science and Technology Council

Department/Unit

Department of Computer Science and Engineering

Website

https://www.grb.gov.tw/search/planDetail?id=13338194

Year

2020

Start date/計畫起

01-08-2020

Expected Completion/計畫迄

31-12-2021

Bugetid/研究經費

684千元

ResearchField/研究領域

資訊科學--軟體

漁業對於台灣而言是個重要的產業，2016年的漁業產值為新台幣860億。因其中44%的產值在於遠洋漁業，故歐盟於2015年給予台灣的黃牌警告讓台灣政府燃起打擊非法漁業(IUU)的決心。如果沒有處理好，歐盟將給予紅牌並禁止台灣漁獲出口而造成約1300萬歐元的損失。除了訂定新的法律"遠洋漁業法"，行政院農委會漁業署也補助發展了新世代海洋監控系統(DeepSea 9)來協助他們監控管理遠洋船隊。雖然這系統已經運作中，但可惜只用來做行政管理。這個計畫將利用已建立的骨幹與資料來進行科學研究。 DeepSea 9系統使用了漁船監控系統(VMS)，但此系統僅限於遠洋船隊。漁業署亦發展了另一套漁船航程記錄器(VDR)來記錄沿近海漁船的航跡。這兩套系統所產生的資料非常巨大，包含了超過58億筆GPS航點，漁獲日誌，轉載與卸魚紀錄，以及其他管理上的資料。我們會研發出新的介面，將兩套系統結合其資料並用於本計畫研究。在這兩套系統的原型設計中，我們發展了高效率的資料儲存方法，可快速地查詢並取回漁船的航跡。前導研究的成果之一，是將這些資料與第三方資訊結合後可產生新的知識。例如我們將漁船的船位資料與氣象資料結合分析後，可以推導出漁港的特性：忙碌漁港、低利用度漁港、以及避風港。基於安全因素，如果低利用度漁港有被當成避風港用途，則不能隨意關閉，並且必須每年給予經費維護。而已經發展成型的DeepSea 9的底層硬體架構與演算法，都是基於這些研究達成的。在本多年期計畫中，除了研究如何合併DeepSea 9與VDR系統之外，我們將發展自動化漁獲熱區分析。我們提出了量化演算法來去除不重要的航點，如漁船處於航行或休息狀態中的航點。我們也將研究如何加速量化的演算法，使他能夠在有效時間內計算並處理巨量的漁業資料。之後透過熱區分析的結果，我們將分析漁船的作業行為與捕獲魚種分布。這階段必須從樣本船戶收集他們的漁獲紀錄與拍賣資料並將這些資料連結到資料庫內，經計算後才能達成。分析出來的結果可以協助我們了解漁獲於環境內的分布，並可讓我們推估整體船隊一年的漁獲總量。最後，我們將利用可處理上百萬資料個體的3D WebGL技術來呈現巨量資料，協助大家了解所有數據的意義。 Fisheries is a vital industry which contributes to Taiwan's economy that produced 86 billion TWD in 2016. Since 44% of the produce is in the overseas fishery sector, the ``yellow card'' sanction warning from the European Commision (EUC) has brought attention to the Taiwan government in fighting illegal, unreported, and unregulated (IUU) fishing. Unless further actions are taken, the EUC will issue a ``red card'' and Taiwan will suffer a loss of 13 million euro yearly for not being able to export marine products. Besides passing new laws, the Fisheries Agency has funded the development of a next-generation oceanic information system (DeepSea 9) to help them monitor, control, and surveil overseas fleet. Although operational, this system is only being used for administrative purposes only. This project will use the available infrastructure and initiate scientific research on the data available. The DeepSea 9 system uses the vessel monitoring system (VMS) system to track vessels, but this is only for overseas fisheries. The Fisheries Agency of Taiwan has another system, namely the voyage data recorder (VDR) system for offshore and coastal fisheries. The information these systems hold is voluminous, including more than 5.8 billion rows of GPS data, fishing logbooks, transshipments and landing records, and administrative information. We will attempt to interface these two systems together and use their data in conjunction for research. The initial design of these systems includes storing the data and provide an efficient way to retrieve voyages of vessels. Preliminary studies show that with the voyages data available in DeepSea 9 and the VDR systems, we aggregate the information with third party weather data to identify roles of fishing ports, i.e., busy ports, unused ports, and sanctuary ports which serves as a hideout for severe weather condition. Moreover, the DeepSea 9 is constructed based on the underlying hardware and algorithms used in the preliminary designs. In the following years, besides attempting to interface the 2 systems into one, we will try to automate the detection of fishing hotspots by using the data in the systems. We propose a quantization algorithm to filter out uninteresting spots, i.e., the fishing vessel is traveling or resting. We will attempt to improve the speed of the algorithm to meet the huge computational demands. Following the hotspot analysis results, we will try to classify fishing activities of vessels, including the method of fishing and the type of fish caught. We will collect logbooks and auction records and correlate them to the database. This will assist us in understanding the distribution of fish schools on the ocean and help us estimate yearly catch. More importantly, we will design visualization systems using 3D WebGL, which is capable of handling more than a million data entities at once, to help people understand what the underlying data means.

Keyword(s)

海洋資訊科學
雲端架構
巨量資料
漁業管理
熱點分析
漁獲推估
作業行為分析
資料視覺化
Oceanic information science
Cloud architecture
Big data
Fisheries management
Hotspot analysis
Fish stock assessment
Fishing activity analysis
Data visualization

DSpace CRIS

Fisheries Data Analysis, Big Data Visulization, and Applications Based on the Next Generation Oceanic Information System.(III)

基本資料

Description