Visualized System Of Location-based Sandstone Landforms

With more than 80 years research of Sandstone Landforms, it has now become a relatively independent branch. However, the current research methods in this field are still traditional, and there is no visualized platform used for data analysis. The combination between information technologies and Sandstone Landforms is highly in demand to achieve faster development. Examples of the problems include nearest Sandstone Landforms recommendation in similar latitude area and massive data analysis, etc.Based on the status quo of current research and combined with the corresponding characteristics, we developed the first visualized platform for structured and unstructured data of Sandstone Landforms. This platform will provide to domestic and foreign scholars to launch related research. Meanwhile, for researcher’s needs of massive pictures collection and classification, we apply location-based service technologies to develop a mobile-side data collection and server-side automatic classification service, which will dramatically reduce the burden of data processing work.Based on general K-D Tree, we proposed an Optimized K-D Tree algorithm to solve the problem of nearest Sandstone Landforms recommendation in similar latitude area. It can greatly reduce candidate points before generating K-D Tree so as to enhance the algorithm efficiency. We also implemented the algorithm and verified the performance. Experiments show that the efficiency is improved by about 23.8% compared to general K-D Tree algorithm. Meanwhile, we conducted data mining work with C4.5, which is a decision tree based algorithm model. We discovered relationships between geological components such as conglomerate, siltstone, claystone and temperature, rainfall and so on. To achieve these algorithms and functions, we built an information system of Sandstone Landforms.Finally, for the design and implementation of the subject, we fully carried out tests and experiments from functionality and performance perspectives. In terms of functionality, all basic functions of information system are running properly, and the current beta version has access to the Internet. In addition, we tested the Optimized K-D Tree algorithm performance and initial start time consumption of mobile-side data collection service, the results proved both performance improvement and actual availability.