Drone For Green - Space Apps Challenge

Team Updates

Trees Height calculation using drone mapping data

Tasnuva Tabassum Oshin

Two Different map analysis Data

Tasnuva Tabassum Oshin

Final map ( 1st mapping )

Tasnuva Tabassum Oshin

Final Map ( 2nd mapping )

Tasnuva Tabassum Oshin

orthomosaic_overlap ( 2nd mapping )

Tasnuva Tabassum Oshin

optimized_camera_position_YZ ( 2nd mapping )

Tasnuva Tabassum Oshin

optimized_camera_position_XZ ( 2nd mapping )

Tasnuva Tabassum Oshin

optimized_camera_position_XY ( 2nd mapping )

Tasnuva Tabassum Oshin

matches_graph_YZ ( 2nd mapping )

Tasnuva Tabassum Oshin

matches_graph_XZ ( 2nd mapping )

Tasnuva Tabassum Oshin

matches_graph_XY ( 2nd mapping )

Tasnuva Tabassum Oshin

geotag_position ( 2nd mapping )

Tasnuva Tabassum Oshin

FC300X_3.6_4000x3000 (RGB)(1)_scaled_pixel_error ( 2nd mapping )

Tasnuva Tabassum Oshin

FC300X_3.6_4000x3000 (RGB)(1)_pixel_error ( 2nd mapping )

Tasnuva Tabassum Oshin

orthomosaic_overlap ( Mapping one )

Tasnuva Tabassum Oshin

optimized_camera_position_XY ( mapping one )

Tasnuva Tabassum Oshin

matches_graph_XY ( mapping one )

Tasnuva Tabassum Oshin

geotag_position ( Mapping one )

Tasnuva Tabassum Oshin

FC300X_3.6_4000x3000 (RGB)(1)_pixel_error ( mapping one )

Tasnuva Tabassum Oshin

dsm_preview ( mapping one )

Tasnuva Tabassum Oshin

texture

Tasnuva Tabassum Oshin

How We did it

https://github.com/Ferdous12/D

Nahid Ferdous

Drone Market Analysis

Nahid Ferdous

Drone For Green ( Data collection and Analysis Part one )

Nahid Ferdous

satellite data analysis Shundorbon , Bangladesh Drone For Green Project

Nahid Ferdous

Full Project summary

“Unmanned Aerial Vehicle based assessment of canopy cover and deforestation dynamics”

Nahid Ferdous

Map Processing

Nahid Ferdous

Concerns and opportunities

•Different forest structures, canopy cover levels, and tree spatial distribution patterns may give the same NDVI value.

•Information from the Bangladesh Forest Department forest inventory plots can be used to evaluate NDVI variability on forested plots.

•UAV imagery can provide accurate and precise estimates of canopy cover over many contiguous Landsat imagery cells.

•Monitoring of spectral change through time likely to improve established relationships between NDVI and forest distribution

Nahid Ferdous

Challenges

•There is no standard NDVI value threshold distinguishing forests from other types of vegetation

•Effect from season progression

•Other spectral metrics, ancillary information, and monitoring of metric values through time for a given location usually very useful

•Corroboration of predictions with ground observations critical

Nahid Ferdous

NDVI-based assessment of vegetation change

•Get Landsat image for area of interest at time T1

•Calculate NDVI (if not already available)

•Repeat for same area at time T2

•Compute new band as NDVIT2 – NDVIT1

•Display the new band

Nahid Ferdous

Normalized

Nahid Ferdous

NDVI

Nahid Ferdous

R code used in the examples

Nahid Ferdous

Several deforestation events in the forest, but substantial reforestation too. Artifacts near the edge of water Deforestation (red) No change (yellow) Reforestation (green)

Nahid Ferdous

More reliable assessment thank to no sensor issues Sporadic deforestation events in the forest, but substantial reforestation too Deforestation (red) No change (yellow) Reforestation (green)

Nahid Ferdous

Lots of artifacts because of 2008 data stripping issue , Deforestation (red) No change (yellow) Reforestation (green

Nahid Ferdous

With masking of cell with value below 0.15 Challenging to quantify challenge with ocular means

Nahid Ferdous

Landsat 8 data Value range [0.00 to 0.86] (more consistent) thanks to new sensor

Nahid Ferdous

Landsat 7, data Stripping issue from sensor degradation evident Value range [-0.31 to 0.82]

Nahid Ferdous

Landsat 7 data, Value range [-0.34 to 0.83] , Minor ocean patterns noticeable

Nahid Ferdous

      
        var landsat = ee.Image('LANDSAT/LE7_L1T_ANNUAL_GREENEST_TOA/2000')
      
        .select(['greenness']);
      
        Export.image.toDrive({
      
        image: landsat,
      
        description: 'NDVI_2000_',
      
        scale: 30,
      
        region: Sundarbans,
      
        maxPixels: 2500000000,
      
        crs: 'EPSG:3106'
      
        });

view raw NDVI Data Analysis ( Shundorbon , Bangladesh ) hosted with ❤ by GitHub

Nahid Ferdous

Space Apps Challenge 2017

Drone For Green | Live Smart

Team Updates

	var landsat = ee.Image('LANDSAT/LE7_L1T_ANNUAL_GREENEST_TOA/2000')

	.select(['greenness']);

	Export.image.toDrive({

	image: landsat,

	description: 'NDVI_2000_',

	scale: 30,

	region: Sundarbans,

	maxPixels: 2500000000,

	crs: 'EPSG:3106'

	});