NM404_AzmuthStar

Size Invariant Ship Detection From SAR Images.

Brief Summary

If you would like to contribute a translation in another language, please feel free! You can add it as a pull request and we will merge it when we get the chance.

This repository is a tutorial for how to use AzmuthStar Apllication on Windows 10, 8, or 7. (It will also work on Linux-based Os with some minor changes.) It was originally written using Python 3.7x, but will also work for newer versions Python.

We have also made a YouTube video that walks through the overview of our project.

This readme describes every step required to run the application :

1. Installation of Gdal
2. Installation of rasterio
3. Installation of Requirements for Python
4. Download Weight File and setup the file path
5. Download Shapefile File and setup the file path
6. Testing

The repository provides all the files needed to run the AzmuthStar Gui Application on your machine and the sample testing image provided in the SAMPLE folder.

Introduction

Problem statement : Size Invariant Ship Detection From SAR Images.

Organisation: Indian Space Research Organisation [ISRO]

Team : AzmuthStar

PSID : NM404

Team_ID: 4141

Team Leader : Krishna Ganpat Kadam

Mentor : Prof. Y. H. Bhosale

Description : SAR satellites provide useful information for object detection. A methodology needs to be developed to detect ships at ocean area using SAR data and estimate size of the detected ships (size invariant). The methodology should be applicable for different resolution SAR data of same bands.

Software Architecture

Software System Requirements

Operating System : Windows 10 , 8 or 7, Linux , Mac.

Steps

1. Installation of Gdal

Gdal (Geospatial Data Abstraction Library) is a useful command tool to process spatial data.

I. Check Python version and MSC version :

1. Download the latest 2.7x version of python or any version you want.
2. install python with the default options and directories.
3. After installation, go to Python IDLE to find out what version of Python you are using and make a note of the number that shows the version of your python in the top right as highlighted below:

I. Install GDAL :

1. Head over to Tamas Szekere's Windows binaries and download the appropriate GDAL binary. For this tutorial we are using MSC v.1900 on 64 bit.the picture below illustrates how to match the version with your own python version.

2. Clicking the link it will take you to the list of binaries (installers) to download.Then Locate the “core” installer, which has most of the components for GDAL.

3. Next, Install the python bindings for your version of Python, this can either be 2.7, 3.1, or 3.x.

4. After downloading the Python bindings, install them.

II. Adding Path Variable :

1. Right click on “Computer” on the desktop and go to “Properties”:
2. Click on Advanced System Properties

3. Select Environment Variables.
4. Under the System variables pane, find the ‘Path’ variable, then click on Edit.

5. Go to the end of the box and copy and paste the following:

C:\Program Files\GDAL

6. In the same System variables pane, click on “New” and then add the following in the dialogue box:

Variable name : GDAL_DATA
Variable value : C:\Program Files\GDAL\gdal-data

7. Click Ok!
8. Add one more new variable by clicking “New…”
9. Add the following in the dialogue box:

Variable name : GDAL_DRIVER_PATH
Variable value : C:\Program Files\GDAL\gdalplugins

III. Testing the GDAL installation :

1. Open CMD and Type:

gdalinfo --version

2. Press Enter.

2. Installation of rasterio

1. Download the rasterio package wheel from Here then,

pip install rasterio-1.1.5-xxx-xxx-win_amd64.whl

3. Installation of Requirements for Python

Clone this repository then go to repo location and hit below command :

pip install -r requirements.txt

4. Download Weight File and setup the file path

1. Yolo_Weights [You can also use your own yolo .weights file & .cfg by replacing them]
2. Download The Yolo Weight HERE.
3. Move .weight and .cfg file in Object_detection/*.

5. Download Shapefile File and setup the file path

1. Shapefile [You can also use your own shapefile for Land Water descrimination: Just go to the Preprocessing.py and paste your shapefile path to LineNo 37.]
2. Download the ShapeFile Here
3. Move Shapefile Directory to processes/

6. Testing

Basically, The testing of GUI is done in the following step :

To initialize the Azmuth Ship Application first of all we need to Login in SAR API by simply clicking on "Download" button and passing Username,password and Product id.

1. Log in SAR api.

2. After pressing download button, the dataset is downloaded under the present working directory of main.py program.
3. Select the Downloaded dataset and it will Unzip in process/Download/Unzip: if we need to unzip the RAW dataset zip file.
4. Now next step is To perform data preprocessing using gdal_translate and gdal_warp on the RAW SAR image , Once the current process completes.
5. We need to proceed for land water descrimination using shape file of ocean, by simply clicking on the BUTTON of "Water Extraction".Along with we are able extract the land portion also.

*Our application will take 3 to 4 minutes to accomplish this process*

6. To achieve the main objective i.e. Ship Detection, Size Estimation from SAR images we need to simply click on the BUTTON of "Detection". Then the below algorithm perform an operations on image.

* Slicing ,Contouring , Detection, Checksum *

7. Then we need to store the coordinate and the Size of the detected ship in the .csv file for plotting the location of ship on world map, Click on "MAP" icon visualize the GEO TAG that specifies the size and length of the ship.

8. Display image.

Done!!