The storage and query of remote sensing image is an important content in geographic information processing and plays an important role in the realtime processing of massive remote sensing images. Aiming at the problem of single node failure, low scalability and low efficiency in traditional remote sensing image processing, this paper proposes a distributed storage and query scheme of remote sensing data based on HBase. In this method, the remote sensing image is divided by uniform mesh, and an index scheme based on grid ID and Hilbert curve is designed according to the block information. Then, the HBase filter mechanism is used to design the filter column family so as to achieve the purpose of screening data in the query. In addition, the MapReduce parallel processing method is adopted to write and query image data. The experimental results show that, compared with MySQL and MapFile, the proposed method can improve the query speed of data and has good scalability.