https://github.com/kevinlacaille/planet_code_challenge_feb_2021
A repo for the code test for the Geospatial Software Engineer position at Planet Labs.
https://github.com/kevinlacaille/planet_code_challenge_feb_2021
Last synced: 7 months ago
JSON representation
A repo for the code test for the Geospatial Software Engineer position at Planet Labs.
- Host: GitHub
- URL: https://github.com/kevinlacaille/planet_code_challenge_feb_2021
- Owner: kevinlacaille
- Created: 2021-02-16T13:25:14.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2021-02-19T20:51:15.000Z (over 4 years ago)
- Last Synced: 2025-01-30T19:28:25.167Z (8 months ago)
- Language: Python
- Size: 14.1 MB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Planet Labs Code Challenge (Feb 2021)
A repo for the code test for the Geospatial Software Engineer position at Planet Labs.
### Summary
This code takes in a time series of Planet Labs' PSScene4Band data, which contains blue, green, red, and near infrared images, and measures how the green vegetation changes over the time series. The measure of greenery is approximated by the measurement of the [normalized difference vegetation
index (NDVI)](https://en.wikipedia.org/wiki/Normalized_difference_vegetation_index) in the scene. See below for additional details.
### Input arguments
```data_directory/```
- The relative or absolute path to your PSScene4Band imagery, for example: ```../task/data/```
- Assumes the following file naming convention for the GeoTIFFs: ```YearMonthDate_XXXXXX_XXXX_3B_AnalyticMS_clip.tif```
- Assumes the following file naming convention for the metadata: ```YearMonthDate_XXXXXX_XXXX_3B_AnalyticMS_metadata_clip.xml```
```output_directory/```
- The relative or absolute path to the program's output figures, for example: ```../outputs/```
### Run the code!
To run this temporal NDVI analysis on your data, type the following into you command line:
```bash
> python3 main.py ../task/data/ ../output/
|████████████████████████████████████████| 7/7 [100%] in 24.9s (0.28/s)
Over the time series, between 0.02% (2020-08-26) and 3.75% (2020-07-01) of the region contained barren dirt.
Over the time series, between 96.25% (2020-07-01) and 99.98% (2020-08-26) of the region contained vegetation.
Average change in NDVI / day: (0.0 +/- 0.12) % per day
Greenness of the vegetation changed over time, however, over the entire time series, the vegetation did not
statistically get greener nor less green over time.
```
### High-level flowchart
Here's how the code works:
1) Parse & validate input arguments
2) Ingest and validate image and metadata files
3) Process images
Extract bands from images → Normalize data → Mask out regions with water → Compute NDVI map → Visualize data → Measure proportions of baren soil and vegetation
4) Compute rate of change of green vegitation compared to baren soil
5) Return how green the vegetation has gotten over time
### Main result
Since the mean change in NDVI over the entire time series was approximetly zero and its standard deviation was >>0, I can conclude that the greenness of
green vegetation changed over time, however, over the entire time series, the vegetation did not statistically get greener nor less green over time.
### Additional details
Please see ```additional_details.md``` for more information about my experience with this challenge.