Baseline drift and other motion artifacts caused by breathing and physical movements need to be reduced from useful signals in pulse wave signals detection process, and the relevant effective methods to calculate pulse rate are also urgently needed. This paper is based on the wearable electronic eyeglasses aiming at measuring pulse signals with a three-axis acceleration sensor to catch body movement signals. It presents a removal method of the baseline drift using cubic spline interpolation, applies the least mean square (LMS) method which needs additional accelerometer signals in reducing the motion artifacts, and adopts the improved exceeding value method to detect the principal peaks and calculate the pulse rates of the pulse signals. Experiment results show that the combination of the three methods can effectively remove the baseline drift, reduce the motion artifacts, detect the principal peaks of pulse signals, and accurately calculate the pulse rates.