Main Article Content
The earth observatory sources generate various types of satellite images with different resolutions such as spatial, spectral, and temporal resolution. Satellite images are very expensive but also there are some freely available satellite images which are used in various applications such as land use land cover classification, agriculture monitoring, fire monitoring, urban monitoring, and flood monitoring, etc. A single satellite generates several terabytes of data per day which means a single satellite creates a huge amount of data to be analyzed. Traditional analysis techniques are not suitable for satellite images because it is less capable to handle large and complex datasets, but big data analysis techniques can be useful to extract meaningful information from the satellite images and it can apply on real-time data as well as offline data. Some of the challenges may occur during the implementation of satellite images with big data analysis such as modeling, processing, mining, querying, and distributing large scales of repositories because it refers to typical datasets which may generate problem during capture, storage, analysis, locating, identifying and securing, and understanding of the data. In this paper, a Feature-Based Image Retrieval (FBIR) system is being developed which will always serve those satellite images which can produce a better result for interpretation and analysis of the earth surface monitoring. The system will always allow retrieving the best possible available image (i.e. less affected image in terms of weather conditions and cloud cover) to the end-user for future use.
This work is licensed under a Creative Commons Attribution 4.0 International License.