本研究提出一種基於中醫脈診特徵與PPG感測技術之低計算資源的個人身分識別系統，目的是提升穿戴式裝置中身份辨識的性能和效率。本研究採用接觸式PPG感測器直接擷取脈搏訊號。為降低記憶體與運算資源的需求，我們引入中醫血液循環共振理論，從PPG訊號中提取12個固定諧振頻率作為脈診特徵，取代傳統依賴時域與頻域複雜特徵擷取的作法。在這基礎上，我們設計一個結合孿生網路與機率式神經網路的混合模型，作為身分識別分類器，以取代常見但計算成本高的CNN分類器。實驗結果顯示，本方法在保有高識別率的同時，顯著減少記憶體使用與推論時間，極具應用於嵌入式穿戴設備之潛力，並開創中醫脈診於現代生物特徵識別領域中的創新應用。;This study proposes a low-resource personal identity recognition system based on Traditional Chinese Medicine (TCM) pulse features and PPG (photoplethysmography) sensing technology, aiming to enhance the performance and efficiency of identity verification in wearable devices. Instead of remote imaging, we adopt a contact-based PPG sensor to directly acquire pulse signals. To reduce memory usage and computational overhead, we introduce the TCM theory of blood circulation resonance and extract 12 fixed resonance frequencies from the PPG signals as pulse features, replacing traditional time-domain and frequency-domain feature extraction methods. Building on this, we design a hybrid model that combines a Siamese network with a Probabilistic Neural Network (PNN) as the identity classifier, in place of conventional yet computationally intensive CNN-based models. Experimental results show that the proposed method maintains high recognition accuracy while significantly reducing memory usage and inference time. This demonstrates strong potential for deployment in embedded wearable devices and represents an innovative integration of TCM pulse diagnosis into modern biometric identification systems
