grhalbgott / index-calculator Goto Github PK

extra argument for range from to (-f 0 -t 0.4)
parse into plot function and give to index_plot

If values are needed for indices, this is easier to keep track of your arguments

like for NDMI, there are no 10m bands, so there has to be help if you want to use the argument 10m

After calculating the index, save some further analytic data into results/ like:

• histogram
• statistics (min, max, mean, std.dev)
• other?

Calculation works, plotting too, but NoData values aren't handled as nan. Why?

Normalized Difference Snow Index
NDSI = (Green - SWIR)/(Green + SWIR)

Riggs, G., D. Hall, and V. Salomonson. "A Snow Index for the Landsat Thematic Mapper and Moderate Resolution Imaging Spectrometer." Geoscience and Remote Sensing Symposium, IGARSS '94, Volume 4: Surface and Atmospheric Remote Sensing: Technologies, Data Analysis, and Interpretation (1994), pp. 1942-1944.

Three things are required:

1. Name of index in full and as abbreviation:

• NBR2
• Normalized Burn Ratio 2

2. Formula (including bands), maybe even reference to according literature

• (SWIR1 – SWIR2) / (SWIR1 + SWIR2)
• Sentinel-2: NBR2 = (B11 – B12) / (B11 + B12)
• Landsat 8-9, NBR2 = (Band 6 – Band 7) / (Band 6 + Band 7)

3. How the plot should look like (output range and cmap)

• RdYlGn
• -1 to 1

the only thing that changes between Landsat and Sentinel is the data structure. Therefore the calculations can be the same but there has to be another variable "data structure/path"

Utils function index_calculator manages then which calculations are possible

Atmospherically Resistant Vegetation Index
ARVI = (NIR – (2 * RED) + BLUE) / (NIR + (2 * RED) + BLUE)

Put all index-calculation in a new file and use utils for:

• read_raster
• cut_raster
• function caller (which function for which index)

Normalized Difference Red Edge Vegetation Index
NDRE = (NIR – RED EDGE) / (NIR + RED EDGE)

Three things are required:

1. Name of index in full and as abbreviation:

• NBR
• Normalized Burn Ratio

2. Formula (including bands), maybe even reference to according literature

• (NIR - SWIR) / (NIR + SWIR)
• Sentinel 2: NBR = (B8A – B12) / (B8A + B12)
• Landsat 8-9, NBR = (Band 5 – Band 7) / (Band 5 + Band 7

3. How the plot should look like (output range and cmap)

• RdYlGn, -1 to 1

Enhanced Vegetation Index
EVI = G * ((NIR – R) / (NIR + C1 * R – C2 * B + L))

https://www.usgs.gov/landsat-missions/landsat-enhanced-vegetation-index

Statistics show wrong values
Histogram seems wrong as well?

https://www.sphinx-doc.org/en/master/tutorial/getting-started.html

Green Normalized Difference Vegetation Index
GNDVI = (NIR – GREEN) / (NIR + GREEN)

OSM-boundaries or gadm.org, check where an API is available and use it to get an aoi (Name of city) if no aoi is specified (if os.path(glob.glob("./data/shapes/*)) not True: get aoi OR if aoi is specified through cmd, get from api

shapefile or geojson

Structure Intensive Pigment Vegetation Index
SIPI = (NIR – BLUE) / (NIR – RED)

https://custom-scripts.sentinel-hub.com/custom-scripts/sentinel-2/indexdb/

Include output options
check for grammar and language

Statistics show wrong values
Histogram seems wrong as well

New mission, same bands, other filenames

Visible Atmospherically Resistant Index
VARI = (Green - Red) / (Green + Red - Blue)

Gitelson, A., et al. "Vegetation and Soil Lines in Visible Spectral Space: A Concept and Technique for Remote Estimation of Vegetation Fraction." International Journal of Remote Sensing 23 (2002): 2537−2562.

Statistics show wrong values
Histogram seems wrong as well

1. write function in indices.py
2. include in index_calculator() in utils.py
3. Check max resolution in resolution_handler() in utils.py
4. define how to plot in plottype_handler() in utils.py
5. Update Readme
6. Commit and close issue

would be like -r resolution
and in glob: "///" + resolution + "/_B0X.jp2"

Use annotate as in random_walker:

plt.annotate(
"Start points: +\nEnd points: x",
xy=(0, 0),
xytext=(5, 5),
xycoords="figure points",
)

Difficulty: has to be projected on the satellite images
simple reprojection won't work

possible with argparse?

call like:
$ python src/main.py roi.shp ndvi ndmi gci vari

cannot show indices one after another, have to save them directly into the folder as pictures

Soil-Adjusted Vegetation Index
SAVI = ((NIR - Red) / (NIR + Red + L)) x (1 + L)
default L = 0,5
but you can put in you own value

In_raster1 = np.array(0, 0, 0, 0, 1, 2, 3)
In_raster2 = np.array(0, 0, 0, 0, 1, 2, 3)

Assert out_raster = zahlen verändert je nach forme

Statistics show wrong values
Histogram seems wrong as well

Is there data you could show? Like differences between two dates or global statistics for the whole file?

Burn Area Index
BAI = 1/((0.1 -RED)^2 + (0.06 - NIR)^2)

Chuvieco, E., M. Pilar Martin, and A. Palacios. "Assessment of Different Spectral Indices in the Red-Near-Infrared Spectral Domain for Burned Land Discrimination." Remote Sensing of Environment 112 (2002): 2381-2396.

Vegetation Condition Index
VCI = (NDVI' - NDVImin) / (NDVImax - NDVImin) * 100

https://iopscience.iop.org/article/10.1088/1755-1315/389/1/012033/pdf

Sentinel-2, Landsat 8 and MODIS/AVHHR are suitable
look up what makes sense next to Sentinel-2 and Landsat

the value range is wrong and the image seems wrong as well

Possible solution:

• calculation is wrong (but it is not, proven by using QGIS Raster Calculator)
• range of values, NoData values

Make the code more readable by changing string format style

Better than including the glob.glob in every index calculation
and change glob.glob to glob (from glob import glob)

Optional argument: save raster
Have to write output file to ./data/ as in clip

Modified Normalized Difference Water Index
MNDWI = (Green − MIR)/(Green + MIR)

Hanqiu Xu (2006) Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery, International Journal of Remote Sensing, 27:14, 3025-3033.

https://iopscience.iop.org/article/10.1088/1755-1315/389/1/012033

-0.8 to 0.8 with three colors (RdYlGn) but with higher contrast (Green - White - Blue) -> viridis?

Green Chlorophyll Vegetation Index
GCI = NIR / GREEN – 1

Normalized Difference Built-up Index
NDBI = (SWIR - NIR) / (SWIR + NIR)

Zha, Y., J. Gao, and S. Ni. "Use of Normalized Difference Built-Up Index in Automatically Mapping Urban Areas from TM Imagery." International Journal of Remote Sensing 24, no. 3 (2003): 583-594.

Use either PyQT or PySimpleGUI
How much work will it be and what is easier?