DEV Community: Kshitij Raj Sharma

Install open source version of VSCode in Ubuntu / Debian

Kshitij Raj Sharma — Wed, 13 Aug 2025 11:38:14 +0000

As a developer, you're likely no stranger to Visual Studio Code (VS Code), a popular, feature-rich code editor developed by Microsoft. However, if you're looking for an open-source alternative that offers similar functionality, look no further than VSCodium.

Why VSCodium?

VSCodium is a community-driven, open-source fork of VS Code. It's designed to be compatible with VS Code extensions and settings, making it an attractive alternative for developers who want to avoid proprietary software.

Installing VSCodium on Ubuntu/Debian

Step 1: Add the VSCodium Repository

First, add the VSCodium repository to your system:

wget -qO - https://gitlab.com/paulcarroty/vscodium-deb-rpm-repo/raw/master/pub.gpg \
    | gpg --dearmor \
    | sudo dd of=/usr/share/keyrings/vscodium-archive-keyring.gpg

Then, add the repository:

echo 'deb [arch=amd64,arm64 signed-by=/usr/share/keyrings/vscodium-archive-keyring.gpg] https://download.vscodium.com/debs vscodium main' \
    | sudo tee /etc/apt/sources.list.d/vscodium.list

Step 2: Update and Install VSCodium

Update your package list and install VSCodium:

sudo apt update && sudo apt install codium

Configuring VSCodium

Once installed, you can configure VSCodium to use the code command, just like VS Code:

sudo ln -s /usr/bin/codium /usr/bin/code

Verify that the code command is working:

code --version

Using VSCodium

You can now launch VSCodium using the code command:

code .

code file.txt

This will open VSCodium with the specified file or directory.

Resources

Known Issues

Some VS Code extensions may not work with VSCodium due to differences in the underlying architecture.
VSCodium may not receive updates as frequently as VS Code.

If you're interested in contributing to the project, check out the VSCodium GitHub repository.

Working with open imagery of the world ? : OpenAerialMap

Kshitij Raj Sharma — Fri, 20 Dec 2024 10:27:37 +0000

Introduction

URL : https://openaerialmap.org

Description : OpenAerialMap is an open service to provide access to a commons of openly licensed imagery and map layer services. Download or contribute imagery to the growing commons of openly licensed imagery. This short tutorial will document the steps needed to process and publish your images in OpenAerialMap

How to upload images ? :

1) Go to openaerialmap
2) Sign in using your google account
3) To upload your image : Hit Upload
4) Provide metadata for your image ( Type of sensor used , License , Creator and etc )
5) Submit , Grab Tea and Wait for the completion
6) Once upload is done you can leave the browser , OAM will do its own postprocessing step and email you when it is done
7) Once image is available in OAM you can share it with your network , Grab it as TMS/ WMTS in your QGIS , Perform mapping using this imagery . Opportunities are unlimited

Troubleshooting

Currently , OpenAerialMap doesn't support more than 4 bands . if your image has more than 4 bands then you need to remove extra bands in order to upload in OAM

Removing extra bands ( Adopt 4 bands limit of OpenAerialMap)

To remove bands / only extract related bands in your image :

For us , We had 6 band image hence we needed to do some preprocessing using QGIS

Reaarange Bands ( GDAL utility )

In QGIS search and find Reaarange bands under GDAL . This function allows you to choose and extract specific bands from
the image and create new tiff with it . We created two orthophotos

Orthophoto with R, G, B and NIR band
Orthophoto with only RGB band

Exporting tiff as rendered image ( fix no display problem / Not rendering issue )

By default with exporting only RGB band as Raw we encountered following issue in OAM ( Image not being rendered ) and not only in OAM its not rendering in ID editor , JOSM and QGIS .

Hence to solve this problem , we exported our RGB tiff as rendered image in QGIS

Black background for nodata problem

Since we used our area of interest to create this orthophoto as there will be pixels with no data in it and by result its being displayed as black in OpenAerialMap and all other places

Hence to solve this problem we clipped our rendrered exported tiff with our Area of Interest in QGIS , ( We had to clip specifically the rendered image output instead of RAW to avoid this problem )

Finally , Now you can upload your rendered clip RGB image to OpenAerialMap.

Resources and Credits

OpenAerialMap
Sujan Paudel
e2e workshop in https://cligeos2024.conscor.com/

Working with Sentinel 1 - RADAR : Case Study in Jajarkot EarthQuake

Kshitij Raj Sharma — Sun, 15 Dec 2024 17:36:43 +0000

The goal of this tutorial is to share my experience and steps I followed when I tried to work with Sentinel-1 Radar Data to estimate the displacement caused by the earthquake in Nepal.

Study Area
Preparation
Download Image
Data Preparation
Image Preprocessing
Generation of Topographic Interferogram
Generation of Differential Interferogram
Creation of Displacement Map
Terrain Correction
Export Displacement Map
Sources and Credit

Study Area

JajarKot Nepal : https://www.openstreetmap.org/relation/4586357#map=10/28.9246/82.3151

Source : https://edition.cnn.com/2023/11/03/asia/nepal-earthquake-northwest-hnk-intl/index.html

There was a deadly earthquake in Early November 2023 which killed more than 157 people as per CNN . I wanted to figure out how much displacement it has cased in the area , Did topography rise ? or fall. As we know due to tectonic plates movement it will always result in displacement !

Preparation

Lets first verify that event exists , get the event details from USGS . USGS mainatains the eartquake catalog which will give the event date, magnitude and coordinate for the event .

https://earthquake.usgs.gov/earthquakes/search/

Prepare your filter accordingly to your event , For us as CNN reported the incident in november 3 , I am keeping filter from Novemeber 2 to 4 in Jajarkot Area . This is how my search filter result looks like : https://earthquake.usgs.gov/earthquakes/map/?currentFeatureId=us7000l8p5&extent=22.2586%2C75.81665&extent=31.72817%2C91.70288&range=search&sort=largest&search=%7B%22name%22%3A%22Search+Results%22%2C%22params%22%3A%7B%22starttime%22%3A%222022-01-01+00%3A00%3A00%22%2C%22endtime%22%3A%222024-12-08+23%3A59%3A59%22%2C%22maxlatitude%22%3A32.319%2C%22minlatitude%22%3A24.121%2C%22maxlongitude%22%3A89.868%2C%22minlongitude%22%3A78.354%2C%22minmagnitude%22%3A5%2C%22orderby%22%3A%22magnitude%22%7D%7D

We found event details , it's coordinate and confirmation from USGS , Now we can move forward to downloading the Sentinel Image

Download Image

Filters

I am using the ASF Data Search by NASA : https://search.asf.alaska.edu/#/

Click on filters Nearby search

Here are the parameters for my Search filters :

Area of Interest : POINT(82.153 28.858) (You can convert your coords to WKT with https://wktmap.com/ )
Start Date : 2023/10/01
End Date : 2023/11/30 (To be on safe side I am trying to find image within two months frame to filter out the best image before and after the event )
File Type : L1 Single Look Complex (SLC) ( SLC has both amplitude and phase , in GRD we have only amplitude . Here in displacement we need phase hence SLC )
Beam Mode : IW ( Because IW- Interforemetric covers Orbit and EW- extra wide swath covers Polar region )

Source : My Lecturer Dr.Karima Hadj-Rabah & Dr.Zahra Dabiri

Polarizations : VV ( Polarization doesn't really matter in phase differences , It matters in classification , Here we are selecting VV just to reduce the image size )
Direction : Ascending ( It doesn't matter as well because at the end we will do geocoding and correct the image so image going to be reotated anyway )

You can get rid of direction , polarization filter if nothing shows up ! It's okay !

Search Results :

I am selecting two image based on my region of interest :

After the event :: Novemeber 10, 2023
Before the event :: October 17,2023

if you look closely to the footprint image top left has my region of interest specifically

Important : You need to make sure both of the image you are downloading ( before and after ) They should be overlapping , otherwise you won't be able to calculate the displacement between two images if they are not of the same region. Which is why for me I get the before and after image of the same footprint to make sure image match and will be easy to visualize later on stacking

You need to create account in order to download the image.

Note :
As some of the required steps are computationally intensive, it is good to store the data at a location which offers good reading and writing speed. If your computer has an internal SSD, processing should be done there to ensure best performance. (Credit : ESA Tutorial)

Data preparation

For data preparation and image processing and analysis I am going to use a tool called SNAP . SNAP (Sentinel Application Platform) is a earth observation processing open-source software developed by ESA

Download SNAP :

https://step.esa.int/main/download/snap-download/

SNAP is available for windows , mac and linux
I will be using linux , I downloaded all tollboxes

Setting up SNAP

chmod +x esa-snap_all_linux-11.0.0.sh

./esa-snap_all_linux-11.0.0.sh

SNAP Operations

Open Snap it , Interface looks like this

Load image

Our downloaded image would be in .zip folder , Don't unzip them simply open them in SNAP

File > Open Product > Select your both before after images > Open

Split data to region of interest

Goto

Radar > Sentinel-1 TOPS > S-1 Tops Split

Navigate to Processing Parameters

Subswath : IW1
Polarisation : VV
Bursts : 6 to 9 (For my area of interest bursts 6 to 9 covered the area , You should select the subswath and bursts accordingly . If you remember my are of intrest was in top left of the image , hence I selected the subswath that covers that area and bursts to minimize it )

Lets rename the output file for readability

Hit Run

You should have the splitted file and now

Visualize the intensity

Expand the image > Bands > subswath name > intensity_* Double click

Now repeat the process to split another image as well , Remember to click on the image that you are splitting or change the path

Visualize the image side by side from

Window > Tile Vertically or horizontally

Image preprocessing

Orbit Correction

Technically we suppose that satellite is following perfect line , But in reality there is vibration in satellite and is not really stable . Hence we need to apply a correction in order to account for the orbital path

Radar > Apply Orbit File

Go to Processing Parameters and Check Do not fail if new orbit file is not found to make sure our process doesn't fail even though orbit file not found ( Because orbit file is only available after two weeks of the acqusition at the moment )

Repeat the process for another image as well . You should have those two new files in your window

Generation of Topographic Interferogram

A topographic interferogram highlights differences in phase caused by terrain elevation. It is a key step in extracting the displacement component. You must align the pixels between two different image ( Before and after one ) to make the comparison hence coregistration

Source ( Dr. Karima )

Cleanup

Lets Clean up the file we don't need

Control select the split file and original image and Right click > Close Products

No need to save Hit NO

Back Geocoding and Enhanced Spectral Diversity

The S-1 Back Geocoding operator coregisters the two split products based on the orbit information added in the previous step and information from a digital elevation model (DEM) which is downloaded by SNAP.

Go to

Radar > Coregistration > S-1 Tops Coregistration > S1- Back Geocoding

Click on Add Opened

Make sure before image is on the top and after image is at the bottom

You should see your image there

Go to Back-Geocoding

Select the DEM source : For me I am selecting SRTM 1Sec

Press Run

You shuold have new stack image with both of your image stacked together , You shouldbe able to see your before and after image intensity

Enhanced Spectral diversity

In order to improve the quality of the after image as related to before , Inorder to remove inospheric error this step is used .

Go to :

Radar > Coregistration > S1 TOPS Coregistration > S-1 Enhanced Spectral Diversity

Hit Run

It should add a file with filename_*stack_esd

You can change the visualization parameter from Color Palette

Interferogram Formation

This is where we actually calculate the phase difference between two images along with coherence

Go To

Radar > Interferometric > Products > Interferogram Formation

if you check on Processing Parameters :

You should see subtract flat-earth phase ticked

Reminder : Make sure input image for the new step is always from the previous step

You should have new filenamewith_ifg extension

TOPSAR Deburst

To remove the demarcation between the single bursts we use TOPSAR Deburst

Go TO :

RADAR > Sentinel 1 - TOPS > S-1 TOPS-Deburst

Click on Run

if you visualize the result , it will remove the small black lines between bursts in the image

Generation of Differential Interferogram

By generating a differential interferogram, we remove the topographic effects, extracting the phase
changes caused by ground displacement.

Source : Dr. Karima

Interferogram formation

Topographic Phase Removal

GOTO :

Radar > Interferometric > Products > Topographic Phase Removal

Navigate to Processing Parameters

Choose the DEM source : For me I chose SRTM 1sec

Check Output Topographic Phase band and the elevation band ( Lets visualize the elevation as well )

Hit Run

I get the output like this along with elevation band

Multilooking

We need to create square pixels to make the resolutions same in range(approx 3m) and azimuth (approx 7m). we are used to see pixels in square

GOTO

RADAR > SAR utilities > Multilooking

Navigate to processing parameters and change number of looks for range and azimuth.

For me I am using

Range Looks : 8
Azimuth Looks : 2

Because it results approx 30m resolution which is enough for me because usualy displacement will happen in larger level, You can see the resolution after the value changes

This is how output looks like

Goldstein Phase Filtering

Lets filter the noise so that we can improve visual interpretability of differential interferogram.

GOTO :

Radar > Interferometric > Filtering > Goldstein Phase Filtering.

Navigate to processing parameters and play with the filter value , For me I am using 0.9 as There is quite some noise in my image

Hit Run

This is what I got as output , I have some noise in the right bottom part of image : for now it is okay but you can also play with it and increase or decrease the filter

Creation of Displacement Map

Dowload snaphu plugin

The wrapped phase values need to be unwrapped to obtain continuous displacement
measurements.

For this output : We need this plugin to be installed : https://step.esa.int/main/snap-supported-plugins/snaphu/

For Debian based distributions, You can install it directly from the apt

sudo apt get update
sudo apt install snaphu

Export env for snaphu

Lets create env needed for snaphu

GO TO
Radar > Interferometric > Unwrapping > SNAPHU Export

Navigate to SnaphuExport

Select the folder where output would be stored ( Recommended to create new folder ) , For me I created snaphu-export folder with snaphu_output filename

Change No of tile rows and tile columns and number of processer based on your laptop spec , For me I used 8 processers and 10 rows/column . Bearing in mind if you set 1 tile per time it will disable the multiprocesser option ! It is advised to process tiles in parallel

Hit Run

Now navigate to the folder you specified you should see following structure

Now open snaphu.conf in one of your editor ( Notepad ++ , VSCode , Kate any you prefer )

Now Here :

Comment out logfile
Find the name of coherence file from the folder and Specify it in CORRFILE

Your final conf should look like this

Now open terminal in the folder dir , Remember in the folder dir

Now you will find the command on top of the file, Copy & run the command in terminal

The time differs based on the performance of laptop & the number of tiles you are processing based on your parameter in snaphu export , For me it took 1 min 43 sec

Import from Snaphu

Now lets import the processing output from snaphu
Goto :

Radar > Interferometric > Unwrapping > Snaphu Import

Select the output file , Remember OUTFILE information is in snaphu.conf

Navigate to : 2 Read-Unwrapped-Phase & Open the result from snaphu

Select .hdr file format

Navigate to Snaphu Import and Tick on Don't save wrapped inferogram in the target Product

Navigate to Write tab

And edit the name add _unwrapped in the name for disntinction

Hit Run

This is how my output looks like

Generate of Displacement Map

This step converts the unwrapped phase into displacement values, enabling the analysis of
ground movement.

Go to

Radar > Interferometric > Products > Phase to Displacement

Hit Run

This is how my displacement map looks like

Terrain Correction

Terrain correction georeferences the displacement map to a coordinate system and removes
distortions caused by topography, ensuring accurate spatial analysis.

Goto :

Radar > Geometric > Terrain Correction > Range-Doppler Terrain Correction

Navigate to Processing Parameters and Select DEM , I selected SRTM 1Sec

Hit Run

This is what I get as output

Lets understand this displacement Map

I changed the visualization from left bottom window and changed editor to slider

If you look into the value , Blue area was the area where terrain went down by approx 16 cm and nearby area terrain raised by 23.7 cm

Export displacement Map

Right click on the image

And Click on Export View as Google Earth KMZ

Lets visualize this KMZ in QGIS

Set layer transparency to 50% so that we can visualize area where terrain went down and terrain went up after earthquake

here is the output and as you can see area is nearly earthquake area and hence we can verify this output makes sense !

You can also export image as geotiff and other formats as well

Thank you for reading !

Sources and Credit :

https://www.youtube.com/watch?v=ZPMRaztNbVU&t=393s
Dr. Karima Hadj-Rabah
My classmate Sahar for her help during blog creation
Paris - Lodron University , Z-GIS Department
ESA Snaphu Tutorial

Process multiband rasters (Sentinel-2) with h3 index and create indices

Kshitij Raj Sharma — Sat, 24 Aug 2024 03:20:10 +0000

Hi, In previous blog we talked about how we can do raster analysis using h3 indexes and postgresql for a single band raster. In this blog we will talk about how we can process multiband raster and create indices with ease. We will be using sentinel-2 image and create NDVI from the processed h3 cells and visualize the results

Download sentinel 2 data

We are downloading the sentinel 2 data from https://apps.sentinel-hub.com/eo-browser/ in Pokhara, Nepal Area , Just to make sure lake is in the image grid so that it would be easy for us to validate the NDVI result

To download sentinel image with all bands :

You need to create an account
Find the image in your area select the grid that covers your Area of interest
Zoom to the grid, And click on icon on right vertical bar
After that go to analytical tab and select all the bands with image format as tiff 32 bit , high resolution , wgs1984 format and all bands checkd

You can also download pregenerated indices such as NDVI , False color tiff only or specific bands whichever best suits your need . We are downloading all the bands as we wanna do the processing by ourselves

Click download

Preprocess

We get all the bands as separate tiff from the sentinel as we downloaded raw format

lets create a composite image :

This can be done through GIS tools or gdal

Using gdal_merge:

We need to rename the downloaded file to band1,band2 like this to avoid slashes in the filename
Lets process upto band 9 for this exercise , you can choose the band as your requirement

gdal_merge.py -separate -o sentinel2_composite.tif band1.tif band2.tif band3.tif band4.tif band5.tif band6.tif band7.tif band8.tif band9.tif

Using QGIS :
Load all individual bands to QGIS
Go to Raster > Miscellanaeous > Merge

While merging you need to make sure you check 'place each input file in sep band'

Now export your merged tiff to raw geotiff as composite
Make sure your image is in WGS1984
in our case image is already is in ws1984 so no need for the conversion
Make sure you don't have any nodata if yes fill them with 0

  gdalwarp -overwrite -dstnodata 0 "$input_file" "${output_file}_nodata.tif"

Finally make sure your ouput image is in COG

  gdal_translate -of COG "$input_file" "$output_file"

I am using the bash script provided in cog2he library to automate those

sudo bash pre.sh sentinel2_composite.tif

Process and creation of h3 cells

Now finally as we have done the preprocessing script , lets move forward to compute h3 cells for each bands in the composite cog image

Install cog2h3

  pip install cog2h3

Export you database credentials

  export DATABASE_URL="postgresql://user:password@host:port/database"

Run We are using resolution 10 for this sentinel image, however you will also see in the script itself which will print the optimal resolution for your raster that makes the h3 cell smaller than your smallest of pixel in raster.

  cog2h3 --cog sentinel2_composite_preprocessed.tif --table sentinel --multiband --res 10

It took a minute for us to compute and store result in postgresql

Logs :

2024-08-24 08:39:43,233 - INFO - Starting processing
2024-08-24 08:39:43,234 - INFO - COG file already exists at sentinel2_composite_preprocessed.tif
2024-08-24 08:39:43,234 - INFO - Processing raster file: sentinel2_composite_preprocessed.tif
2024-08-24 08:39:43,864 - INFO - Determined Min fitting H3 resolution for band 1: 11
2024-08-24 08:39:43,865 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:44,037 - INFO - Resampling Done for band 1
2024-08-24 08:39:44,037 - INFO - New Native H3 resolution for band 1: 10
2024-08-24 08:39:44,738 - INFO - Calculation done for res:10 band:1
2024-08-24 08:39:44,749 - INFO - Determined Min fitting H3 resolution for band 2: 11
2024-08-24 08:39:44,749 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:44,757 - INFO - Resampling Done for band 2
2024-08-24 08:39:44,757 - INFO - New Native H3 resolution for band 2: 10
2024-08-24 08:39:45,359 - INFO - Calculation done for res:10 band:2
2024-08-24 08:39:45,366 - INFO - Determined Min fitting H3 resolution for band 3: 11
2024-08-24 08:39:45,366 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:45,374 - INFO - Resampling Done for band 3
2024-08-24 08:39:45,374 - INFO - New Native H3 resolution for band 3: 10
2024-08-24 08:39:45,986 - INFO - Calculation done for res:10 band:3
2024-08-24 08:39:45,994 - INFO - Determined Min fitting H3 resolution for band 4: 11
2024-08-24 08:39:45,994 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:46,003 - INFO - Resampling Done for band 4
2024-08-24 08:39:46,003 - INFO - New Native H3 resolution for band 4: 10
2024-08-24 08:39:46,605 - INFO - Calculation done for res:10 band:4
2024-08-24 08:39:46,612 - INFO - Determined Min fitting H3 resolution for band 5: 11
2024-08-24 08:39:46,612 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:46,619 - INFO - Resampling Done for band 5
2024-08-24 08:39:46,619 - INFO - New Native H3 resolution for band 5: 10
2024-08-24 08:39:47,223 - INFO - Calculation done for res:10 band:5
2024-08-24 08:39:47,230 - INFO - Determined Min fitting H3 resolution for band 6: 11
2024-08-24 08:39:47,230 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:47,239 - INFO - Resampling Done for band 6
2024-08-24 08:39:47,239 - INFO - New Native H3 resolution for band 6: 10
2024-08-24 08:39:47,829 - INFO - Calculation done for res:10 band:6
2024-08-24 08:39:47,837 - INFO - Determined Min fitting H3 resolution for band 7: 11
2024-08-24 08:39:47,837 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:47,845 - INFO - Resampling Done for band 7
2024-08-24 08:39:47,845 - INFO - New Native H3 resolution for band 7: 10
2024-08-24 08:39:48,445 - INFO - Calculation done for res:10 band:7
2024-08-24 08:39:48,453 - INFO - Determined Min fitting H3 resolution for band 8: 11
2024-08-24 08:39:48,453 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:48,461 - INFO - Resampling Done for band 8
2024-08-24 08:39:48,461 - INFO - New Native H3 resolution for band 8: 10
2024-08-24 08:39:49,046 - INFO - Calculation done for res:10 band:8
2024-08-24 08:39:49,054 - INFO - Determined Min fitting H3 resolution for band 9: 11
2024-08-24 08:39:49,054 - INFO - Resampling original raster to: 200.786148m
2024-08-24 08:39:49,062 - INFO - Resampling Done for band 9
2024-08-24 08:39:49,063 - INFO - New Native H3 resolution for band 9: 10
2024-08-24 08:39:49,647 - INFO - Calculation done for res:10 band:9
2024-08-24 08:39:51,435 - INFO - Converting H3 indices to hex strings
2024-08-24 08:39:51,906 - INFO - Overall raster calculation done in 8 seconds
2024-08-24 08:39:51,906 - INFO - Creating or replacing table sentinel in database
2024-08-24 08:40:03,153 - INFO - Table sentinel created or updated successfully in 11.25 seconds.
2024-08-24 08:40:03,360 - INFO - Processing completed

Analyze

Since now we have our data in postgresql , Lets do some analysis

Verify we have all the bands we processed ( Remember we processed from band 1 to 9 )

select *
from sentinel

Compute ndvi for each cell

explain analyze 
select h3_ix , (band8-band4)/(band8+band4) as ndvi
from public.sentinel

Query Plan :

QUERY PLAN                                                                                                       |
-----------------------------------------------------------------------------------------------------------------+
Seq Scan on sentinel  (cost=0.00..28475.41 rows=923509 width=16) (actual time=0.014..155.049 rows=923509 loops=1)|
Planning Time: 0.080 ms                                                                                          |
Execution Time: 183.764 ms                                                                                       |

As you can see here for all the rows in that area the calculation is instant . This is true for all other indices and you can compute complex indices join with other tables using the h3_ix primary key and derive meaningful result out of it without worrying as postgresql is capable of handling complex queries and table join.

Visualize and verification

Lets visualize and verify if the computed indices are true

Create table ( for visualizing in QGIS )

create table ndvi_sentinel
as(
select h3_ix , (band8-band4)/(band8+band4) as ndvi
from public.sentinel )

Lets add geometry to visualize the h3 cells This is only necessary to visualize in QGIS , if you build an minimal API by yourself you don't need this as you can construct geometry directly from query

ALTER TABLE ndvi_sentinel  
ADD COLUMN geometry geometry(Polygon, 4326) 
GENERATED ALWAYS AS (h3_cell_to_boundary_geometry(h3_ix)) STORED;

Create index on geometry

create index on ndvi_sentinel(geometry);

Connect your database in QGIS and visualize the table on the basis of ndvi value Lets get the area near Fewa lake or cloud

As we know value between -1.0 to 0.1 should represent Deep water or dense clouds
lets see if thats true ( making first category as transparent to see the underlying image )

Check clouds :

Check Lake

As there were clouds around the lake hence nearby fields are covered by cloud which makes sense

Thank you for reading ! See you in next blog

Raster Analysis Using Uber h3 indexes and PostgreSQL

Kshitij Raj Sharma — Mon, 12 Aug 2024 13:51:37 +0000

Hi , In this blog we will talk about how we can do Raster analysis with ease using h3 indexes and as a bonus we will also talk about how it can be served as vector tiles and display on the map

Objective

For learning, We will do calculation on figuring out how many buildings are there in settlement area determined by ESRI Land Cover. Lets aim of national level data for both vector and raster .

Let's first find the data

Download Raster Data

I have downloaded the settlement area from Esri Land Cover .

https://livingatlas.arcgis.com/landcover/

Lets download the 2023 year , of size approx 362MB

Download OSM Buildings of Nepal

Source : http://download.geofabrik.de/asia/nepal.html

wget http://download.geofabrik.de/asia/nepal-latest.osm.pbf

Preprocess the data

Lets apply some preprocessing to data before actual h3 cell calculations
We will be using gdal commandline program for this step. Install gdal in your machine

Conversion to Cloud Optimized Geotiff

If you are unaware of cog , Checkout here : https://www.cogeo.org/

Check if gdal_translate is available

gdal_translate --version

It should print the gdal version you are using

Reproject raster to 4326

Your raster might have different source srs , change it accordingly

gdalwarp -overwrite esri-settlement-area-kathmandu-grid.tif esri-landcover-4326.tif -s_srs EPSG:32645 -t_srs EPSG:4326

Important : Incase your raster also has nodata , We need to convert it to 0 using -dstnodata 0 to make sure our h3 cells are not being built on nodata area

gdalwarp -overwrite esri-settlement-area-kathmandu-grid.tif esri-landcover-4326.tif -s_srs EPSG:32645 -t_srs EPSG:4326 -dstnodata 0

Now lets convert tif to cloud optimized geotif

gdal_translate -of COG esri-landcover-4326.tif esri-landcover-cog.tif

It took approx a minute to convert reprojected tiff to geotiff

Insert osm data to postgresql table

We are using osm2pgsql to insert osm data to our table
& downloaded pbf from geofabrik

osm2pgsql --create nepal-latest.osm.pbf -U postgres -E 4326

osm2pgsql took 274s (4m 34s) overall.

You can use geojson files also if you have any using ogr2ogr

ogr2ogr -f PostgreSQL  PG:"dbname=postgres user=postgres password=postgres" buildings_polygons_geojson.geojson -nln buildings

ogro2gr has wide range of support for drivers so you are pretty flexible about what your input is . Output is Postgresql table

Calculate h3 indexes

Postgresql

Install

pip install pgxnclient cmake
pgxn install h3

Create extension in your db

create extension h3;
create extension h3_postgis CASCADE;

Now lets create the buildings table

CREATE TABLE buildings (
  id SERIAL PRIMARY KEY,
  osm_id BIGINT,
  building VARCHAR,
  geometry GEOMETRY(Polygon, 4326)
);

Insert data to table

INSERT INTO buildings (osm_id, building, geometry)
SELECT osm_id, building, way
FROM planet_osm_polygon pop
WHERE building IS NOT NULL;

Log and timing :

Updated Rows    8048542
Query   INSERT INTO buildings (osm_id, building, geometry)
    SELECT osm_id, building, way
    FROM planet_osm_polygon pop
    WHERE building IS NOT NULL
Start time  Mon Aug 12 08:23:30 NPT 2024
Finish time Mon Aug 12 08:24:25 NPT 2024

Now lets calculate the h3 indexes for those buildings using centroid . Here 8 is h3 resolution I am working on . Learn more about resolutions here

ALTER TABLE buildings ADD COLUMN h3_index h3index GENERATED ALWAYS AS (h3_lat_lng_to_cell(ST_Centroid(geometry), 8)) STORED;

Raster operations

Install

pip install h3 h3ronpy rasterio asyncio asyncpg aiohttp pandas

Make sure reprojected cog is in static/

mv esri-landcover-cog.tif ./static/

Run script provided in repo to create h3 cells from raster . I am resampling by mode method : This depends upon type of data you have . For categorical data mode fits better. Learn more about resampling methods here

Install

pip install cog2h3

python cog2h3.py --cog esri-landcover-cog.tif --table esri_landcover --res 8 --sample_by mode

or using commandline directly

cog2h3 --cog esri-landcover-cog.tif --table esri_landcover --res 8 --sample_by mode

Log :

2024-08-12 08:55:27,163 - INFO - Starting processing
2024-08-12 08:55:27,164 - INFO - COG file already exists: static/esri-landcover-cog.tif
2024-08-12 08:55:27,164 - INFO - Processing raster file: static/esri-landcover-cog.tif
2024-08-12 08:55:41,664 - INFO - Determined Min fitting H3 resolution: 13
2024-08-12 08:55:41,664 - INFO - Resampling original raster to : 1406.475763m
2024-08-12 08:55:41,829 - INFO - Resampling Done
2024-08-12 08:55:41,831 - INFO - New Native H3 resolution: 8
2024-08-12 08:55:41,967 - INFO - Converting H3 indices to hex strings
2024-08-12 08:55:42,252 - INFO - Raster calculation done in 15 seconds
2024-08-12 08:55:42,252 - INFO - Creating or replacing table esri_landcover in database
2024-08-12 08:55:46,104 - INFO - Table esri_landcover created or updated successfully in 3.85 seconds.
2024-08-12 08:55:46,155 - INFO - Processing completed

Analysis

Lets create a function to get_h3_indexes in a polygon

CREATE OR REPLACE FUNCTION get_h3_indexes(shape geometry, res integer)
  RETURNS h3index[] AS $$
DECLARE
  h3_indexes h3index[];
BEGIN
  SELECT ARRAY(
    SELECT h3_polygon_to_cells(shape, res)
  ) INTO h3_indexes;

  RETURN h3_indexes;
END;
$$ LANGUAGE plpgsql IMMUTABLE;

Lets get all those buildings which are identified as built area in our area of interest

WITH t1 AS (
  SELECT *
  FROM esri_landcover el
  WHERE h3_ix = ANY (
    get_h3_indexes(
      ST_GeomFromGeoJSON('{
        "coordinates": [
          [
            [83.72922006065477, 28.395029869336483],
            [83.72922006065477, 28.037312312532066],
            [84.2367635433626, 28.037312312532066],
            [84.2367635433626, 28.395029869336483],
            [83.72922006065477, 28.395029869336483]
          ]
        ],
        "type": "Polygon"
      }'), 8
    )
  ) AND cell_value = 7
)
SELECT *
FROM buildings bl
JOIN t1 ON bl.h3_ix = t1.h3_ix;

Query Plan :

This can further be enhanced if added index on h3_ix column of buildings

create index on buildings(h3_ix);

When shooting count : there were 24416 buildings in my area with built class classified as from ESRI

Verification

Lets verify if the output is true : Lets get the buildings as geojson

WITH t1 AS (
  SELECT *
  FROM esri_landcover el
  WHERE h3_ix = ANY (
    get_h3_indexes(
      ST_GeomFromGeoJSON('{
        "coordinates": [
          [
            [83.72922006065477, 28.395029869336483],
            [83.72922006065477, 28.037312312532066],
            [84.2367635433626, 28.037312312532066],
            [84.2367635433626, 28.395029869336483],
            [83.72922006065477, 28.395029869336483]
          ]
        ],
        "type": "Polygon"
      }'), 8
    )
  ) AND cell_value = 7
)
SELECT jsonb_build_object(
  'type', 'FeatureCollection',
  'features', jsonb_agg(ST_AsGeoJSON(bl.*)::jsonb)
)
FROM buildings bl
JOIN t1 ON bl.h3_ix = t1.h3_ix;

Lets get h3 cells too

with t1 as (
  SELECT *, h3_cell_to_boundary_geometry(h3_ix)
  FROM esri_landcover el
  WHERE h3_ix = ANY (
    get_h3_indexes(
      ST_GeomFromGeoJSON('{
        "coordinates": [
          [
            [83.72922006065477, 28.395029869336483],
            [83.72922006065477, 28.037312312532066],
            [84.2367635433626, 28.037312312532066],
            [84.2367635433626, 28.395029869336483],
            [83.72922006065477, 28.395029869336483]
          ]
        ],
        "type": "Polygon"
      }'), 8
    )
  ) AND cell_value = 7
)
SELECT jsonb_build_object(
  'type', 'FeatureCollection',
  'features', jsonb_agg(ST_AsGeoJSON(t1.*)::jsonb)
)
FROM t1

Accuracy can be increased after increasing h3 resolution and also will depend on input and resampling technique

Cleanup

Drop the tables we don't need

drop table planet_osm_line;
drop table planet_osm_point;
drop table planet_osm_polygon;
drop table planet_osm_roads;
drop table osm2pgsql_properties;

Optional - Vector Tiles

To visualize the tiles lets quickly build vector tiles using pg_tileserv

Download pg_tileserv Download from above provided link or use docker
Export credentials

export DATABASE_URL=postgresql://postgres:postgres@localhost:5432/postgres

Grant our table select permission

GRANT SELECT ON buildings to postgres;
GRANT SELECT ON esri_landcover to postgres;

Lets create geometry on h3 indexes for visualization ( You can do this directly from query if you are building tiles from st_asmvt manually)

ALTER TABLE esri_landcover 
ADD COLUMN geometry geometry(Polygon, 4326) 
GENERATED ALWAYS AS (h3_cell_to_boundary_geometry(h3_ix)) STORED;

Create index on that h3 geom to visualize it effectively

CREATE INDEX idx_esri_landcover_geometry 
ON esri_landcover 
USING GIST (geometry);

  ./pg_tileserv

Now you can visualize those MVT tiles based on tiles value or any way you want eg : maplibre ! You can also visualize the processed result or combine with other datasets. This is the visualization I did on h3 indexes based on their cell_value in QGIS

References :

Exploring Overture Map Data

Kshitij Raj Sharma — Sun, 09 Jun 2024 01:22:55 +0000

Welcome ! First lets start with overture , if you don't know what is Overture Maps Foundation and what it does I strongly recommend you to go through this website : https://overturemaps.org/ , I tried to build small utilities and hosted them so that readers of this blog also can look into the data and analyze by themselves.

Release Used

Overture release: 2024-05-16-beta.0

Objectives

Primary Objective

To perform qualitative and quantitative analysis of Overture map data.

Secondary Objectives

Visualize the releases on a country level.
Conduct qualitative analysis to identify additions to existing OSM data and differences across countries.
Facilitate general users in forming their own opinions based on the available data.

Approach

Build a script to retrieve Overture data as geoparquet with multiple themes (streamlining and automating the process).
Convert geoparquet to geojson.
Convert flattened geojson to pmtiles.
Develop a viewer for comparison and loading.
Automate the entire process with a bash script.
Compare with population data, existing OSM buildings in the area, and if possible, the number of people per building.

Considerations

Duckdb, overturemaps-py, and GDAL were tested for extraction, with overturemaps-py standing out as simple and perfect. The repo was forked, and enhancements were added to the viewer and filters to support any custom key and value.
Tippecanoe was used to convert geojsonseq to pmtiles.
A bash script was used to automate the entire process, making it configurable using config.json (base and default) for layers, their properties, tile generation settings, combining multiple layers into a single tile, and fetching the right key and value for specific layers.
The primary statement being validated is: "Overture Maps data will undergo validation checks to detect map errors, breakage, and vandalism to help ensure that map data can be used in production systems."

Study Areas

Argentina
Indonesia & Malaysia Area
Kenya
Liberia
Malawi
Nepal
Nigeria

Note: Covering bounding boxes were drawn to somehow match the country boundary in above listed countries ( this is not true for all of them - actual boundary may differ ). Data on those bbox were downloaded, viewed, analyzed, and compared regarding its distribution and how it fits with the existing population.

View Geojson Here

Qualitative Analysis

Buildings

Buildings seem to have undergone good conflation.
Offset and merging of ML datasets have been taken care of.
Buildings present on satellite images seem to be included in the dataset.

Roads

Roads are not cleaned and validated.

When a release is published, there are no major enhancements, and orphan roads remain in the datasets.
Tags are not fixed or validated (For eg: In Nepal, most of the roads were classified as unclassified - same as OSM. Some major roads have inconsistency in trunk and primary). It appears that tags validation is still ongoing or something is not being looked into.

Some Validation Issues

Pular Pisau, Borneo (Near Malaysia):

Height feature is present in only some buildings. In countries like Nepal, it is minimal.

Inconsistent tags in road dataset along with orphan roads as mentioned above
Meanwhile , POI datasets appear to be detailed and populated in most places, making them easily importable into OSM. You need to be aware of confidence though

Quick summary

Overture datasets stand out well for building footprints and POIs, relatively speaking. Transportation, Land, and Land Use seem somewhat similar to OpenStreetMap. (This is before overture released new land cover datasets which I haven't looked into)
Validation and conflation are poor in layers other than buildings.
Good offset alignment with roads.

Quantitative Analysis

Buildings

Area	Google Open Buildings	%	Microsoft ML Buildings	%	OpenStreetMap (as per Overture info)	%	Total Overture Buildings	Population Estimate	P.E. (in mil)	People per Building	Approx Current OSM Buildings
Argentina	34,545,592	73%	8,998,855	19%	3,457,499	7%	47,001,946	78,765,589	78.77	1.68	3,497,866
Liberia	1,557,014	55%	144,185	5%	1,148,863	40%	2,850,062	10,157,546	10.16	3.56	1,151,027
Indonesia	4,314,085	41%	2,485,377	24%	3,641,263	35%	10,440,725	27,523,228	27.52	2.64	3,651,924
Nepal	26,280,737	68%	4,396,928	11%	8,078,311	21%	38,755,976	129,874,888	129.87	3.35	8,243,272
Malawi	8,882,648	61%	1,758,044	12%	3,927,989	27%	14,568,681	29,256,446	29.26	2.01	3,943,125
Kenya	20,334,091	59%	3,734,399	11%	10,414,457	30%	34,482,947	75,320,339	75.32	2.18	10,557,014
Nigeria	50,787,453	68%	7,150,013	10%	16,304,722	22%	74,242,188	252,698,591	252.70	3.40	17,966,401

Overture release: 2024-05-16-beta.0

PS: Population and Current OSM Buildings Estimate is from Kontour API

People per building = Population Estimate on the Area / Total Overture Buildings
Approx current OSM buildings = Fetched from the OSM at current date to validate the overture osm building numbers may not match as overture kept snapshot of osm and by the time of this analysis buildings might increase or decrease in osm, should give rough idea

Analysis was not done on exact country boundary, its bbox taken in the area as provided in the geojson and shared the same geometry using different parameters

Places distribution based on confidence value

According to overture confidence values in places is about the existence of the place itself, which means if it has 50 % that means there is 50/50 chance that place exists there. I tried to see how much can I trust may be those above 80 % ? or 70 so I tried to figure out how much data is there in this threshold .

Country	Above 90 % Confidence	80-90 %	70-80 %	50-70 %	Below 50 %
Argentina	0.438	17.3557	1.6333	38.1136	42.4594
Indonesia & Malaysia Area	0.1412	12.3793	0.3198	47.8856	39.2741
Kenya	0.2197	12.8847	1.8883	41.023	43.9842
Liberia	0	10.0957	0.3299	58.1326	31.4418
Malawi	0.1422	12.9801	1.2269	51.0135	34.6373
Nepal	0.4004	11.0466	5.9221	33.3404	49.2904
Nigeria	0.1078	10.2943	1.3312	38.5526	49.7141
Average	0.2070428571	12.43377143	1.807357143	44.00875714	41.54304286

P.S. Table is in percentage distribution for example in Argentina out of POI available there only 0.4 percentage of data with more than 90 % confidence

Conclusion

From the qualitative analysis conducted on different parts of the world, the data is impressive in terms of offset management when different sources are grouped. I am preetty amazed to see the coverage along with conflation and offset accross the different parts of the world. Buildings seem to be well-matched with each other on an obsolete level, and when ground truth checking with Esri imagery, it covers most places. However, when combined with the tabular analysis in most of the places people-per-building ratio are not that realistic yet they are not worst too (seems it doesn't left out and covers most , it might have some extra clutter buildings). For example, in Argentina, it's 1.68 which seems pretty low. It appears that OpenStreetmap buildings are preserved and are as told (given highest priority - if you look into current approx OSM buildings and numbers included in overture they are quite similar). A massive number of AI building footprints are added to the datasets, whereas google buildings are almost more than 50% in all of the area (Except Indonesia). For roads, validation is still poor, especially in areas like Nepal and Indonesia, where many orphan roads exist in the datasets.I expect tags validation and cleaning specially on road which is not case in the areas I looked into , tags such as primary roads , trunk , unclassified roads are inconsistent. The POI datasets seem well-detailed, and there is great potential for them to be added to OSM after validation, as RapID already has this functionality. While doing so you need to be aware that higher confidence data is low as compared to number of datasets available . On average : only 0.2 % are of above 90 % and 12.4 % on 80 - 90 percent confidence values so even though total row numbers are large better to filter them based on higher confidence. 3D height data is not impressive in the developing countries yet I was surprised to see some of them in countries like Nepal. Building footprints seems to be well defined and aligned with transportation layers exploring the potential that it might be quickly checked validated and used in case of pre-disaster response.

This analysis might be incomplete and is my only personal view with quick analysis on the area I looked into. It is suggested to form your own opinion using the developed tools and data shared as shown in the video by the end of this blog.

Tools and Resources Developed

Querier

https://queryparquet.streamlit.app/ (Tool might go in sleep mode if there is no usage , Please wake it up if needed)

Source code : https://github.com/kshitijrajsharma/qrp

*Features : *

Allows you to shoot custom queries on the parquet data , such as stats , how many rows are their which are from microsoft , meta etc
Default query to get stats based on the source
Provides a box where you can form your own query if you like
Integrates OpenStreetMap current buildings and population of the area (based on bbox supplied) so that you can use it in your query for the analysis
Supports remote parquet url as input and prepopulates the study area that I did

Example dirty query to get % distribution for places

Viewer

I made a quick dirty viewer to do qualitative analysis, The viewer can directly be accessed from Querier or also available here: https://hotosm.github.io/overture-to-tiles/

Viewer supports remote pmtiles and custom styling , Example viewer with default styling : https://hotosm.github.io/overture-to-tiles/?url=https%3A%2F%2Fstaging-raw-data-api.s3.amazonaws.com%2Fdefault%2Foverture%2F2024-05-16-beta.0%2Fargentina%2Fpmtiles

Source code : https://github.com/kshitijrajsharma/overture-to-tiles/tree/master/docs

*Viewer features : *

Simultaneously view the place in OpenStreetMap , Overture and ESRI satellite image
Open all layers in OpenStreetMap Editor : RapID
Allows user to download geoparquet of source
Query the attributes and tile bounds
Custom styling supports for the vector layers like this : https://github.com/kshitijrajsharma/overture-to-tiles/blob/master/docs/styles/default.js
Supports remote pmtiles using url parameter
Toggle vector layers and their classes along with OpenStreetMap and ESRI Satellite image
3D view with both overture height data and OSM no of floors data

Quickly View Study Area Datasets

Extractor

https://github.com/kshitijrajsharma/overture-to-tiles/blob/master/scripts/Readme.md

https://github.com/kshitijrajsharma/overture-to-tiles/

*Extractor features : *

Automates extraction from overture data using custom theme : https://github.com/kshitijrajsharma/overture-to-tiles/blob/master/scripts/base_theme.json
Supports children layers to be combined into single pmtiles layer
s3 upload

Quick demo how you can visualize and analyze the data

Watch Video :

https://github.com/kshitijrajsharma/overture-data-analysis-report/assets/36752999/31eb9917-3fff-42db-9f5d-d2c53649bb81

Resources and Credits :

Pmtiles , Overture-py , Tippecanoe , Overture-docs , RapID

I welcome your thoughts and comments.

Install latest gdal in debian

Kshitij Raj Sharma — Tue, 14 May 2024 13:37:48 +0000

Hi , I am writting short steps about how to install latest gdal version in your debian stable.

By the time I am writting this post the default version of gdal available in the stable debian bookworm is 3.6.5 but I want the latest one . Usually gdal will publish the latest release in the unstable source first. if you wanna get your hand on it either you need to install it from the source or you can install it from the unstable source. I will list how to install from unstable source only for the gdal

1) First update sour package list



 sudo nano /etc/apt/sources.list

Add following line



deb http://deb.debian.org/debian/ unstable main contrib non-free

2) Update package list



sudo apt update

3) Now install gdal-bin using unstable source



sudo apt install -t unstable gdal-bin

4) Check your gdal version



ogrinfo --version

Hurray !

Install gdal python

if you want python installation of gdal then you can follow following steps :

Install gdal dev libraries ```bash

sudo apt-get install libgdal-dev

- Export env variables 
```bash


export CPLUS_INCLUDE_PATH=/usr/include/gdal
export C_INCLUDE_PATH=/usr/include/gdal

Now install gdal using pip ```bash

pip install gdal==

You are all set !

Share Postgresql from Windows to WSL Linux

Kshitij Raj Sharma — Tue, 14 May 2024 12:02:33 +0000

Hi , I am writting this short tutorial about how to share your current postgresql with postgis extension installed in your windows system with WSL. We will use networking share method.
For me , I needed to Install postgersql on my harddisk and I had it already before installing WSL and in need of sharing . I am documenting the steps you can take if you wanna download and install it in windows

Install and remember your postgresql and postgis version

I followed graphical installation method to install postgis postgresql bundle in my windows
Download the Installer :
https://www.enterprisedb.com/downloads/postgres-postgresql-downloads
Installing postgresql 15 in my case

Install Postgis extension from stack builder

You can follow this awesome tutorial in case you get lost

Once you have postgresql on your windows up and running , Now we need to access it from the WSL we have

Install postgresql client in your wsl

sudo apt install poststgresql-client

Lets configure network mode in wslconfig

In your windows terminal (Inside Default profile : for eg : In my case C:\Users\mouse)

wsl nano .wslconfig

[wsl2]
networkingMode=mirrored

Check the connection

 psql -h localhost -p 5432 -U postgres -d postgres

Restart WSL

wsl --shutdown

Now you should be able to connect to your windows localhost from your WSL.

Monitor GPU Usage in WSL debian

Kshitij Raj Sharma — Tue, 14 May 2024 08:52:46 +0000

NVTOP is like htop but for your graphics module . In this short tutorial I will share how to install nvtop in wsl debian

OS used :

cat /etc/os-release



PRETTY_NAME="Debian GNU/Linux 12 (bookworm)"
NAME="Debian GNU/Linux"
VERSION_ID="12"
VERSION="12 (bookworm)"
VERSION_CODENAME=bookworm
ID=debian
HOME_URL="https://www.debian.org/"
SUPPORT_URL="https://www.debian.org/support"
BUG_REPORT_URL="https://bugs.debian.org/"

1) Verify your graphics driver
In my case I have nvidida graphics



nvidia-smi

It should display details about your graphics , if not first install it

2) Edit your debian source list to include non-free releases



sudo nano /etc/apt/sources.list

2) Modify the source list and add following



deb http://deb.debian.org/debian/ bookworm main contrib non-free-firmware

3) Get update



sudo apt get update

4) Now you can install nvtop



sudo apt install nvtop

Now hit nvtop and you are all set !!

Install Debian in Different Location -WSL Windows

Kshitij Raj Sharma — Tue, 14 May 2024 08:17:16 +0000

I was trying to install Debian in different location rather than the C drive . Sharing the steps I took .

OS : Windows 11
WSL Version : Wsl 2

1) First install Debian as by default in C

wsl --install -d Debian

2) Shut down wsl

wsl --shutdown

3) Export Debian to your desired location

In my case its E :

 wsl --export Debian E:\wsl_export\debian-ex.tar

4) Now unregister your old Debian from wsl

wsl --unregister Debian

5) Now import Debian to the Disk

wsl --import Debian "E:\wsl\Debian" "E:\wsl_export\debian-ex.tar"

6) Cleanup
You can remove your backup now and set Debian as default

wsl --set-default Debian

Extra tips

To make your user in debian superuser : After migrating you might loose your default login using username you created before .

sudo usermod -aG sudo username

Change default user

nano /etc/wsl.conf

Replace username with your username

[user]
default=username

Now wsl --shutdown and login debian again you should be logged in as your default user with sudo access

DEV Community: Kshitij Raj Sharma

Install open source version of VSCode in Ubuntu / Debian

Why VSCodium?

Installing VSCodium on Ubuntu/Debian

Step 1: Add the VSCodium Repository

Step 2: Update and Install VSCodium

Configuring VSCodium

Using VSCodium

Resources

Known Issues

Working with open imagery of the world ? : OpenAerialMap

Introduction

How to upload images ? :

Troubleshooting

Removing extra bands ( Adopt 4 bands limit of OpenAerialMap)

Exporting tiff as rendered image ( fix no display problem / Not rendering issue )

Black background for nodata problem

Resources and Credits

Working with Sentinel 1 - RADAR : Case Study in Jajarkot EarthQuake

Table of Contents

Study Area

Preparation

Download Image

Filters

Data preparation

Download SNAP :

SNAP Operations

Load image

Split data to region of interest

Image preprocessing

Orbit Correction

Generation of Topographic Interferogram

Cleanup

Back Geocoding and Enhanced Spectral Diversity

Enhanced Spectral diversity

Interferogram Formation

TOPSAR Deburst

Generation of Differential Interferogram

Interferogram formation

Multilooking

Goldstein Phase Filtering

Creation of Displacement Map

Dowload snaphu plugin

Export env for snaphu

Import from Snaphu

Generate of Displacement Map

Terrain Correction

Export displacement Map

Sources and Credit :

Process multiband rasters (Sentinel-2) with h3 index and create indices

Download sentinel 2 data

Preprocess

Process and creation of h3 cells

Analyze

Visualize and verification

Raster Analysis Using Uber h3 indexes and PostgreSQL

Objective

Let's first find the data

Download Raster Data

Download OSM Buildings of Nepal

Preprocess the data

Conversion to Cloud Optimized Geotiff

Insert osm data to postgresql table

Calculate h3 indexes

Postgresql

Raster operations

Analysis

Verification

Cleanup

Optional - Vector Tiles

References :

Exploring Overture Map Data

Release Used

Objectives

Primary Objective

Secondary Objectives

Approach

Considerations

Study Areas

Qualitative Analysis