心跳資訊是臨床檢測生命參數的重要指標，現行的主流方法是採用接觸式檢測技術，非接觸式(Non-contact)的檢測是指借助外來能量，不接觸人體的方式，來檢測人體生理環境的變化。 目前一般使用於量測身體狀況參數的生理監視儀器，皆須要以有線的方式，將感測器聯結於量測者的身體上，長時間下來對量測者而言是相當不舒服的。非接觸式偵測的設備由於價格過於昂貴及取得不易，不適合在一般家庭或是個人使用。 本研究提出一套基於影像之非接觸式心跳量測方法，利用人臉對於光線吸收的變化來找出心跳，並過影像處理與獨立成份分析法將影像中隱含的資訊找出並集中，再透過頻域轉換及心跳估測等程序，達到全自動非接觸量測。 Clinical testing heart beat information is an important indicator of life parameters; the current mainstream approach is the use of contact detection technology. Non-contact detection refers to the use of external energy without contacting the human body to detect changes of the human body in the physiological environment. Physiological monitor instrument is generally connecting the human body by using wired sensor to measure human body’s physical parameters; it is uncomfortable for people to using this monitoring instrument for a long time. Non-contact detection devices are not suitable for use in general household or personal, because its price is too expensive and difficult to acquire. This study proposes a non-contact method based on video format to measure the heartbeat. It uses the information of the absorption amount of our face to various lights to predict the heartbeat; it is also through image processing and Independent Components Analysis (ICA) to find the hidden information in images and to integrate these information, and then through frequency domain transformation and heartbeat estimation procedure, to achieve a complete automatic non-contact measurement.
