An integration of multi-physiological signals for developing emotion recognition engine system is provided, including: an input unit of physiological signals, for sensing concurrent various multi-physiological signals, wherein the various multi-physiological signals include at least a facial expression signal, the facial expression signal is obtained through following steps: recoding plural coordinate positions via a facial image defining a plurality of distance data (Di) via an equation: normalizing the distance data through a normalizing equation: selecting main portions of the normalized distance data calculating supporting vectors with the main portions through multi-class SVM, relevance SVM and fuzzy SVM setting up comparison vectors and identifying items with a training database, the supporting vectors recognized in comparing with the comparison vectors and identified with the identifying items producing an emotional signal via the corresponding identifying item an integration unit of physiological signals, for receiving multi-physiological signals from the input unit, computing and integrating each of the physiological signals, wherein the computation and integration of each of the physiological signals is to integrate physiological signal values computed from the concurrent physiological signals into an emotion status database a signal feature selecting unit, for selecting important physiological signals via feature selection algorithm according to the emotion status database after integrating an emotion computation and recognition unit, for computing and recognizing corresponding emotion data according to the important physiological signals selected from the signal feature selecting unit an output unit, for outputting an emotion code according to the emotion data to a device to show up an emotion status corresponding to the emotion code. An integration of multi-physiological signals for developing emotion recognition method is also provided, using an inte
