計畫團隊成員

 
圖片1-2

總計畫暨子計畫一
Main Project & Subproject 1

蕭勝夫  教授

國立中山大學資工系

 
圖片2-2

子計畫二
Subproject 2

張雲南  副教授

國立中山大學資工系

圖片3-2

子計畫三
Subproject 3

陳坤志  助理教授

國立中山大學資工系

圖片4-2

子計畫四
Subproject 4

陳嘉平  教授

國立中山大學資工系

 

技術亮點 Technical Highlights


本團隊運用了人工智慧的尖端科技,並實際應用於水產養殖之監控,能透過水下鏡頭及各式感測器的輔助下,讓我們在原本混濁的水下能夠拍攝出清晰的影像畫面,並利用機器學習的技術來辨識出魚蝦及餌料。此外,可依據各種環境參數(水質混濁程度、溫度等),動態調整嵌入在感測裝置的神經網路模型硬體加速器運算相關參數,達到低功耗省電的即時魚蝦和餌料辨識偵測運算系統,並可根據運算結果,自動開啟智慧監控系統之其他周邊控制設備(如增氧機、自動餵餌機等)。

This project applies cutting-edge artificial intelligence technology to the smart aquaculture monitor system . With the aid of underwater lenses and various sensors and using machine learning techniques, we are able to generate clear images under the original turbid water and identify fish, shrimp, and bait in acqua-culture ponds. In addition, according to various environmental parameters (remained battery energy, water turbidity, temperature, etc.), the neural network hardware accelerator embedded in the edge devices (camera/sensors) can be dynamically adjusted to achieve low power and real-time processing speed for fish/shrimp/bait detection in the underwater monitor system. Furthermore, the computation results can be used to activate the peripheral control equipment in the smart underwater aqua-culture system.

 

應用情境 Applications


本技術是要建構一個邊際運算之智慧型水下養殖監控系統,提高業者產能、降低成本。此系統是由一個太陽能供電之整合型感測模組組成,包括水下攝影機和各種感測器系統,能即時執行水下影像強化和蝦/餌料殘餘量之偵測,將養殖池和魚蝦的各種現場狀況(如魚蝦及餌料的比例、水質等),呈現在養殖業者之手持裝置上,而且處理後的結果可用以自動啟動智慧養殖監控系統的其他周邊控制設備(如自動餵餌機、增氧機、注水/排水系統等)。
This technology is to construct a smart underwater breeding monitoring system for edge computing, which will increase the productivity of the aqua-culture industry and reduce costs. The system consists of a solar-powered integrated sensing module that includes underwater cameras and various sensors that real-time perform underwater image enhancement and detection of shrimp/bait residuals. The various on-site environment conditions (such as the proportion of fish and shrimp and bait, water quality, etc.) are presented on the handheld device of the operator, and the processing results can be used to automatically activate other peripheral control devices (such as automatic feeding machine, aerators, water injection/drainage systems, etc.).
 

示意圖 Schematic Diagram