Avinash Nunna 20161176 Nikhilendra Atheti 20161054
This project presents a technique to estimate the amount of green cover and urban cover when a satellite image is given. The proposed method is based on preprocessing the satellite image for image enhancement and use NDVI values along with the conventional RGB channels to get a segmentation technique which is able to classify green cover, urban cover and water cover. Using the two classified outputs collected in the classification step, a much better and accurate classification is obtained. The implementation is fully unsupervised and is used to generate the classification of the LANDSAT-8 satellite image.
1)color_image_processing.py outputs the enhanced image. 2)the above run to pan_sharp.py to get the pansharped image. 3)ndvi_calculation.py outputs the landsat ndvi 4)the folder src has segmnetation codes which output the Final result.
link to input images- https://drive.google.com/open?id=11FNj5cPnBAitDH_ZhIn-KNcLoE-VJAJA link to output images - https://drive.google.com/open?id=1J_mu81S2pHK_1QYQRCchCliCwaputiVC