Data aggregation technique removes redundant message in original data, reduces the amount of data transmission, reduces network communication overhead, improves message acquisition accuracy and collection efficiency, and prolongs the effective working time of the network. It is an energy-saving technique widely used in wireless sensor networks, one of the techniques. However, with the continuous operation in the wireless sensor network, the data aggregation technique saves the energy cost of the network, but also brings new security risks to the network. Therefore, how to ensure the confidentiality and integrity of data in the process of aggregation, calculation, storage, and forwarding is the core link of wireless sensor network security research. This paper uses SURF algorithm to study this question. The algorithm in this paper not only has good robustness to the data, but also introduces data aggregation and box filter in the calculation, so that the registration time can be better improved. After the research in this paper, the accuracy of the algorithm in this paper is 19% better than the traditional algorithm. At the same time, the accuracy of the aggregated data increases over time as the channel congestion decreases in the traditional process.