LARGE SCALE SATELLITE IMAGE PROCESSING USING HADOOP FRAMEWORK

Computer Engineering

The processing of large scale of images is needed when there are satellite images involved. Now a day's amount of data continues to increase as more information becomes available. And it will increase amount of surface and recognition, segmentation, and event detection in satellite images with a highly scalable system becomes more and more desirable. In this paper, a construction of semantic taxonomy for the land-cover classification of satellite images. Overall system is constructed in a Distributed HADOOP Computing platform. This system is divided into two important part Training and Running classifier. The Training part classifies images one after another such as Vegetation, Building, Pavement, water, snow etc. Large files are distributed and further divided into multiple data nodes. The map processing jobs placed on all nodes are operated on their local copies of the data. It can be observed that the name node stores only the metadata and the log information during the data transfer to and from the HDFS is done through the Hadoop API. Training classifier is constructed using HADOOP MapReduce Framework and is based upon Google Earth. The Running classifier performs zoom-in, zoom-out and calculates the difference between old and new images.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.