In large-scale Wireless Sensor Networks (WSNs), with limited computing power and storage capacity of sensor nodes, there is an urgent demand of high performance sensory data processing. This study studies the interconnection of wireless sensor networks and cloud-based storage and computing infrastructure. It proposes the idea of distributed databases to store sensory data and MapReduce programming model for large-scale sensory data parallel processing. In our prototype of large-scale sensory data processing system, Hadoop Distributed File System (HDFS) and HBase are used for sensory data storage, and Hadoop MapReduce is used for data processing application execution framework. The design and implementation of this system are described in detail. The simulation of environment temperature surveillance application is used to verify the feasibility and reasonableness of the system, which also proves that it significantly improves the data processing capability of WSNs.