A HBase Based Massive Remote Sensing Metadata Search System

As one of integrated earth observation technologies, Remote sensing technology has play a more and more important role in the fields of military, agriculture, mining, marine and so on. For the past few years, Remote sensing satellite technology in China has achieve a huge improvement, the size of remote sensing images has been changed from the TB level to PB level, what’s more, data types are also tend to be more and more diverse. There are increasing demands from different areas for storage and efficient retrieval of remote sensing data. The metadata of remote sensing images plays an important role in remote sensing applications. It has the characteristic of complex structure, large amount of information and generated with high frequency. Store and retrieve the metadata in traditional relational database will confront with read and write performance issues, poor scalability problems when with deal with huge amount of metadata.We designed and developed a distributed remote sensing metadata retrieval system based on HBase and Elasticsearch named Zodic, which combines the No SQL technologies and the research project in our laboratory. Due to the inability of query functions in HBase, our system implemented a SQL-like query engine based on HBase which support efficient keyword query, aggregate query and spatial query. Also our system support the full text search of remote sensing metadata based on distributed search engine Elasticsearch. According to the experiment result, our system has a high performance, great scalability, thus it can be well applied to the retrieval of massive remote sensing metadata.