Aerial-Satellite-Imagery-Retrieval

1. Chirag Khandhar
2. Akshay Kulkarni
3. Megha Tatti

The Bing Map Tile System

• We are using Bing Map Tile System that provides a pre render World Map at multiple levels of details.
• We provide a lat-lon bounding box and Level of detail to cut each map into tiles for quick retrieval.
• To make the map seamless, and to ensure that aerial images from different sources line up properly, we have to use a single projection for the entire world.
• We chose to use the Mercator Projection which looks like this:

Ground Resolution and Map Scale

• In addition to the projection, the ground resolution or map scale must be specified in order to render a map.
  • Lowest Level (Level 1) = map size 512 x 512 px
  • Level 2 = map size = 1024 x 1024 px and so on.
  • Thus, the map width and height grow by a factor of 2.
• In general, we calculate the width and height of the map in pixels as follows:
  • Map width = Map Height = 256 x 2^level pixels
• The Ground Resolution indicates the distance on the ground that’s represented by a single pixel in the map.
• It varies depending on the level of detail and the latitude at which it’s measured.
• Thus, by using Earth Radius = 6378137 m the Ground Resolution in m/px can be calculated as follows:
  • Ground Resolution = cos(latitude * pi/180) * earth circumference / map width
  • Ground Resolution = (cos(latitude * pi/180) * 2 * pi * 6378137 meters) / (256 * 2^level pixels)
• The Map Scale indicates the ratio between map distance and ground distance, when measured in the same units.
• It varies depending on the level of detail and the latitude at which it’s measured.
• It can be calculated from the ground resolution as follows, given the screen resolution in dots per inch, typically 96 dpi:

  • Map Scale = 1 : ground resolution * screen dpi / 0.0254 meters/inch

  • Map Scale = 1 : [cos(latitude * pi/180) * 2 * pi * 6378137 * screen dpi] / (256 * 2^level * 0.0254)

Pixel Coordinates

• Now after calculating the above quantities, we now convert the Geographic Coordinates into Pixel Coordinates.
• We consider the following conventions:
  • Pixel at upper left corner = (0, 0)
  • Pixel at lower right corner = (width - 1, height - 1)
• Given latitude and longitude in degrees, and the level of detail, the pixel XY coordinates can be calculated as follows:
  • sinLatitude = sin(latitude * pi/180)
  • pixelX = ((longitude + 180) / 360) * 256 * 2 level
  • pixelY = (0.5 – log((1 + sinLatitude) / (1 – sinLatitude)) / (4 * pi)) * 256 * 2^level

Tile Coordinates and Quadkeys

• To optimize the performance of map retrieval and display, the rendered map is cut into tiles of 256 x 256 pixels each.
• Each tile is given XY coordinates from upper left to lower right corner.
• Thus, given a pair of pixel XY coordinate we can easily determine the tile XY coordinates of the tile containing that pixel.
  • tileX = floor( pixelX / 256)
  • tileY = floor( pixelY / 256)
• The quad key is a integer value with base 4 that is accepted by the Bing map.
• Parameters for Quadkey:
  • Tile Position: A tuple of tile coordinates x and y.
  • Level: The level of detail of the map ranging from 1 to 23 that was used to calculate the pixel position.

How to Run

• This script was written in Python 3.8.
• Install the following packages
  • numpy
  • cv2
  • requests
• We've used Anaconda IDE (Spyder).
• Run the main.py by pressing run button or if you are using Command Prompt use python main.py
• Enter the required data.
  • p1_latitude = 49.945895
  • p1_longitude = 7.846655
  • p2_latitude = 49.952333
  • p2_longitude = 7.820331
  • level = 16
  • filename = "Test.jpg"