An adaptive measurement method is proposed based on Android mobile terminals streaming video services QoE, which is adapted to solve the problem of difficulty and inaccuracy in getting quality of experience(QoE) data under the different wireless environments (3G, WiFi). By measuring and evaluating QoE of online video service in real time, it can be more accurate and practical. Thus, an application that can automatically measure QoE of videos is designed to measure objective quality of service (QoS) parameters and map to the subjective QoE (MOS value) through a utility function. Finally, we improve the theoretical model by studying the correlation between theoretical QoE assessment model (taken from the literature) and actual user feedback. It shows that the results obtained by the improved model measuring, under wireless environment is closer to the actual user feedback, which can better evaluate the QoE.