Panda Aerospace

https://docs.google.com/presentation/d/1hIk8vvfLIxpRQP05OGLQNNR30s_dRdvVHLTYUkB-TNU/edit?usp=sharing

https://github.com/Yologuy101/Panda-Aerospace

Our project mainly focuses on using satellite imagery to help with urban planning/development. This is done through image processing of certain areas to determine where urban planners can build roads. Currently our app accepts an image of a map, then an edge detection algorithm is applied or a vegetation detection algorithm is applied. The edge detection algorithm we used is Canny. After applying this algorithm, the image being processed can be easily read to determine where an existing road is and where an open area is. The masking algorithm uses CIELAB (LAB) color space, and by setting the “a” variable to ~0, we were able to detect all the green vegetation and draw them on an output image. This is helpful to the urban designer because it will allow them to determine where vegetation is and map out areas of forest, parks, and residential areas. If an area is forest parks or residential areas, no roads should be built. If an area has moderate vegetation, then it is ok to build roads. To develop this application, we used Python for the scripts, Jupyter Notebook for testing, libraries such as opencv, matplotlib, numpy, PIL (Pillow) and a computer to code on. By making this program, we hope to provide the urban planners with helpful insight to aid their decision making/design decisions.

For SUDIS we mainly used Google World/Map and Landsat. We also used Nasa’s Popgrid viewer and Alos Global Digital Surface as inspiration. For Google Maps and Landsat, we used it primarily to get test map examples for our code run through. However, even though we didn’t use the Nasa Popgrid viewer or Alos GLobal Digital Surface for any part of our program, it inspired us to create the “Yellow” filter idea that would identify below sea level areas and potentially identify dense population areas for SUDIS.

During the brainstorming period, we narrowed down our options to either “COVID-19: Calculate the Risk” or “Drones and Satellites for Urban Development.” We ultimately decided on the Urban Development project because we felt it would better reflect our strengths. Overall, we would describe our learning experience through this hackathon as fruitful and we learned a lot on how to delegate tasks to different members of the team, and how to ask for help whenever we felt a bit overwhelmed, or didn’t understand what we saw when researching. We approached this program from the ground up. First, we came up with some questions and researched how UAVs and satellites could help in the mapping of urban landscapes and urban development. Then, we wrote a pseudocode that we could use as a base for what we wanted to accomplish in this hackathon. As for the actual coding, we started with edge detection to map the streets, and also added a layer for the vegetation in the area. We encountered a few challenges along the way, but they were pretty simple to get through because of good communication. We collaborated and talked through each problem, and we asked our mentors if we had any questions. We would like to thank our mentors Dr. Gao and Mr. Lin for helping us throughout the project and keeping our morale high when we were frustrated. We would also like to thank Ms. Cathy for organizing this entire event and getting everyone together.

• SeDAC POPGRID Viewer https://sedac.ciesin.columbia.edu/mapping/popgrid/ 
• Alos GLobal Digital Surface https://www.eorc.jaxa.jp/ALOS/en/aw3d30/index.htm
• Eoscdis worldview https://worldview.earthdata.nasa.gov 
• https://d-nb.info/118793013X/34
• https://www.sciencedirect.com/science/article/pii/S0886779815302923
• https://rheologic.net/articles/urban-wind-assessment/
• https://www.epa.gov/green-infrastructure/reduce-urban-heat-island-effect
• https://www.satimagingcorp.com/applications/environmental-impact-studies/land-cover-and-change-detection/

#Panda

This project has been submitted for consideration during the Judging process.