智慧邊際運算之物件行為辨識系統
本系統基於智慧視覺感測與無失真壓縮技術,來建構出一個智慧生活與居家生活系統。主要技術是使用高效能影像分析技術,透過簡易可更新背景模型來得到良好的前景資料與抗環境干擾性。在無失真壓縮技術方面,主要是考量IoT環境的有限傳輸頻寬,同時又要兼具Bio-signal高信賴度的特性而發展的。目前主要考量心電訊號監護儀器及多導程的心電圖量測裝置,透過即時且高壓縮率的訊號壓縮處理方法,可避免龐大資料量造成網路頻寬與儲存空間的負擔。基於上述技術,目前已成功開發出數個可實體展示的應用。
Abstract:
This system is designed for intelligent life in smart home. The technology is based on intelligent computer vision technology and lossless data compression. It utilizes the simple adaptive background model to get good foreground and anti-environmental interference. Additionally, the real-time lossless compression is designed not only to improve diagnostic accuracy but also for subsequent medical applications. It mainly concerns the limited bandwidth issue on IoT environment. The designed is implemented on ECG monitor system without any loss on signal. Based on the above technology, some demonstration systems are derived.







