Analyzing massive databases of many years worth of Sat. Images is labor intensive and extremely difficult by hand. Researchers, scientists, and policy makers do not have the ability to get the data they need from Satellite images

Solution: Develop software and utilize algorithms that have the abilities to “weed” out the majority of what you DON’T want and/or provide a small percentage of what you DO want

This program utilizes a Python script to retrieve and sift through specific spatial and temporal data. Once the data is retrieved we utilized Cvision Consulting LLC’s Image Annotator Software we were able to develop a JSON file and CSV file that annotates where specific objects of interest are located within an image for further analysis and characterization.

This could be utilized for any types of objects.

OpenCV no longer includes the SURF and SIFT algorithms which made our planned analysis a little more difficult. The JP2 format the images come in are large and fairly difficult to work with in an efficient manner.

Within the 2 days we had to work we believe we developed a great starting point for a powerful piece of software, that has the potential to save thousands of hours of manual searching through images.

Share this project:

Updates