To reduce the data redundancy and energy consumption caused by the explosive growth of data in wireless sensor networks, this paper presents a node data image based mean filtering algorithm. In the provided approach, the nodes are divided into active and de-active nodes to eliminate the data of the sleep nodes and data redundancy. The nodes of wireless sensor network are clustered, and the data obtained from the sensor network are modeled. Simulation results show that the proposed algorithm is more effective than that of k-medoids, which can reduce the redundant data. Moreover, it reduces the energy consumption, and improves the life cycle of wireless sensor networks. DOI: http://dx.doi.org/10.5755/j01.itc.47.3.20565