Size-Invariant-Ship-Detection-from-SAR-Images

1. Processing geo tiff files in python:

		- Read and write geo tiff files
	
2. Remove Speckle noise using NDSAR filter.


3. Generate Superpixels using fast-slic.


4. Use K-means to cluster the superpixels.


5. Identify and remove the land-superpixels.


6. Use GSHHG data to remove the landmass overlooked by Superpixel-merging.


7. Train VGG16 faster RCNN model for SHIP DETECTION.


8. Return the count of Ships.


9. Estimate Ship size using Haversine formula.


10. Output a Vector file with:

		- Location(XY - Latitude and Longitude).
		- Ship Count.
		- Ship Size.

11. Create a platform independent application for all the functional aspects to converge.
		
		-Currently looking into electron.js