課程編碼
Course Code
中文課程名稱
Course Name (Chinese)
英文課程名稱
Course Name (English)
總學分數
Credits
總時數
Hours
6505169 衛星遙測酬載之Edge-AI實作 Edge-AI Implementation for Satellite Remote Sensing Payloads 3.0 3
中文概述
Chinese Description
本課程聚焦合成孔徑雷達(SAR)與邊緣人工智慧(Edge-AI),結合微波遙測、雷達訊號處理與嵌入式實作。內容涵蓋電磁波與極化、地表散射、雷達方程式、SAR 觀測幾何、距離與方位解析度、脈衝壓縮、合成孔徑聚焦及實測影像處理。學生使用 Python 模擬回波並完成距離壓縮與二維成像,再以實際 SAR 影像建立分類或分割模型,於 Jetson AGX Orin、Orin Nano 及 i.MX8M Plus 進行部署、量化與效能評估。AI GPU 工作站負責模擬、資料處理與訓練,邊緣平台驗證影像判讀、感興趣區域保留及資料減量。課程採十六加二週;前十六週每週一小時理論、兩小時實作,第十五、十六週為期末專題報告;第十七、十八週為自主學習,每週三小時。
英文概述
English Description
This course integrates synthetic aperture radar (SAR) with edge artificial intelligence (Edge-AI) through microwave remote sensing, signal processing, and embedded laboratories. Topics include polarization, surface scattering, the radar equation, SAR geometry, range and azimuth resolution, pulse compression, coherent focusing, and measured SAR image processing. Students simulate radar echoes in Python, implement range compression and two-dimensional imaging, and train classification or segmentation models with measured SAR imagery. An AI GPU workstation supports simulation and training; Jetson AGX Orin, Orin Nano, and i.MX8M Plus support deployment, quantization, and evaluation for image interpretation, region selection, and data reduction. The 16+2-week course provides one lecture hour and two laboratory hours per week during weeks 1–16. Weeks 15–16 are reserved for final project presentations and demonstrations. Weeks 17–18 provide three hours of self-directed learning per week.
核心能力指標 4.設計工程系統、元件或製程之能力。
6.發掘、分析、應用研究成果及兼顧永續發展,以解決複雜且整合性工程問題的能力。
7.認識時事議題,瞭解工程技術對環境永續、社會共好及全球發展的影響,並培養持續學習的習慣及能力。

備註: