Mapping our land more clearly and colourfully
for better analysis and visualisation

Acknowledging Landsat

Celebrating Landsat's 40th

NASA: Landsat

USGS: Landsat

 

We at GeoSage wish to help the public realise the full potential of Landsat, the most well-known and popular satellite. It's
YOUR satellite
and highly complementary to the high-resolution imaging!

Our aim is to develop the highest-quality Landsat-8 imagery composite at 15m resolution.

 

Some user comments:

"Your software made my job easy." (Email from a user, 28 June 2013)

"Thank you so much for your time and great tools." (Email from a user, 8 July 2013)

"Thanks for such an amazing and quick little product!" (Email from a user, 9 August 2013)

"I tried out Spectral Transformer for Landsat-8. Worked a treat. Thanks for making this free." (Twitter from a user, 20 August 2013)

"I've managed to compose bands and merge with pan. It's amazing!" (Email from a user, 20 August 2013)

"We ran across GeoSage's recent press release for the Spectral Transformer Landsat-8 software, downloaded the software and tried it out. It is a very impressive piece of software and created very crisp and natural looking images." (Email from an industry veteran, 22 October 2013)

"I am using your Spectral Transformer and all works perfectly... Many thanks to you for this great and accurate tool." (Email from a user, 16 November 2013)

...

"This is a great tool. I've already tried it with landsat 8 and with just a few clicks, the results are impressive. Thanks for sharing." (LinkedIn user comment in the RSPSoc group, 7 August 2014)

"I enjoy your GUI tool. It's like an App on PC and the first button click acts as simple as the 'voice' input in Google search." (Comment from an industry veteran, 8 August 2014)

Related ...

World basemaps


global basemaps

 

Rapid response to major events: Case studies

9 March 2014: Help search for the missing Malaysia Airlines Flight 370 with the latest Landsat-8 satellite imagery: info download updated on 10 March 2014.

 

 

 

 

 

 

 


 

 

 

 






 

(1) Spectral Transformer for Landsat-8 imagery: Free, advanced image stretching and pan-sharpening software tool to easily and rapidly make colourful and detail-rich (15m-resolution) Landsat-8 imagery composite (News release 17 June 2013, PDF )

20 July 2013: More than 1,500 downloads during the first month of release
22 July 2014: GUI tool released - 2 simple button clicks for fast band combinations, image stretching and pan-sharpening of Landsat-8 imagery ("Perfect tool to take the daily fresh Landsat-8 imagery into GIS and web mapping")
3 November 2014: Added a new button for easy conversion from stretched or pan-sharpened results in GeoTIFF to Google Earth KMZ format (link to third-party free software, see a tutorial below)

Earth Observation (EO) is the most effective way to monitor the environment of this home planet. The latest significant contribution by the U.S. government is the successful lunch and operation of the Landsat-8 satellite. Its high-quality and rich multispectral data are freely available and provide enormous potential for innovation and applications.

Landsat 8
Landsat-8 captures about 400 scenes per day. The U.S. Geological Survey (USGS)  distributes Landsat-8 data in three very accessible ways:

The USGS Landsat portal also provides comprehensive FAQs in relation to the new Landsat-8 imagery and its comparison with the previous Landsat series. It is important to read these before conducting proper image processing. Landsat-8 products are delivered as 16-bit images with the panchromatic band at 15m resolution and multispectral bands at 30m resolution, and band combinations are unique (e.g. bands 4/3/2 refer to red/green/blue, respectively).

An overview of some common band combinations for better discriminating various ground features is provided here, there.


New analysis tools


While the imagery source is magnificent, more work needs to be progressed on how to use the imagery in a straightforward way.

  • There are many remote sensing and image processing software tools on the market, but it is fair to say that very few can efficiently make beautiful, detail-rich imagery composites with adaptive image histogram stretching and advanced image pan-sharpening. One may spend hours to produce something that is of high quality.
  • For many GIS users, it is often hard to find right capable tools (i.e. band combination, image stretching and image pan-sharpening) in GIS software packages.
  • And for casual users and the general public, dedicated tools to process the vast Landsat imagery archive are lacking.
Spectral Transformer tools for Landsat-8 imagery fill in this gap.

Standalone tool set performs three steps of analyses:
    • Step 1 - Band combination (to make three-band imagery composite)
    • Step 2 - Image histogram stretching (to make colourful composite)
    • Step 3 - Image pan-sharpening (to make spatially sharper and colourful composite)
We believe these tools are very useful for a wide range of users who are interested in analysing the Landsat-8 imagery.

(2) Three DOS-based command-line tools for rapidly making colourful Landsat-8 imagery composites

These powerful and easy-to-use tools specifically target Landsat-8 imagery in GeoTIFF format directly downloaded from the USGS Landsat-8 distribution portals, e.g. GloVis.

Step 1: Simple and convenient band combination to make three-band imagery composite

Landsat8_RGB.exe
<red_band_in> <green_band_in> <blue_band_in> <RGB_out.tif>

e.g. Landsat8_RGB.exe b4.tif b3.tif b2.tif RGB_b432.tif
(for true colour)

e.g. Landsat8_RGB.exe b5.tif b4.tif b3.tif CIR_b543.tif
(for colour infrared)
 
e.g. Landsat8_RGB.exe b7.tif b5.tif b3.tif FC_b753.tif
(for false colour)

Step 2: Adaptive linear and non-linear image histogram stretching over the three-band imagery composite

Landsat8_Stretch.exe
<red_band_order> <green_band_order> <blue_band_order> -s
<image_stretch_band1_left_cut> <image_stretch_band1_right_cut> <image_stretch_band2_left_cut> <image_stretch_band2_right_cut> <image_stretch_band3_left_cut> <image_stretch_band3_right_cut> <nonlinear_stretch_band1>
<nonlinear_stretch_band2>
<nonlinear_stretch_band3>
<RGB_in.tif> <RGB_stretched_out.tif>

e.g. Landsat8_Stretch.exe 1 2 3 -s 2 0.2 2 0.2 2 0.2 2 2 2
RGB_b432.tif RGB_b432_stretched_res30m.tif
e.g. Landsat8_Stretch.exe 1 2 3 -s 2 0.2 2 0.2 2 0.2 0 0 0
RGB_b432.tif RGB_b432_stretched_res30m.tif

e.g. Landsat8_Stretch.exe 1 2 3 -s 0 0 0 0 0 0 0 0 0
RGB_b432.tif RGB_b432_stretched_res30m.tif

Step 3: Advanced and fast image pan-sharpening

Landsat8_Fuse.exe
<pan-sharpening_method> -s
<image_stretch_band1_left_cut> <image_stretch_band1_right_cut> <image_stretch_band2_left_cut> <image_stretch_band2_right_cut> <image_stretch_band3_left_cut> <image_stretch_band3_right_cut> <nonlinear_stretch_band1>
<nonlinear_stretch_band2>

<nonlinear_stretch_band3>
<pan_band_in.tif> <ms_in.tif> <fused_out.tif>

e.g. Landsat8_Fuse.exe 1 -s 2 0.2 2 0.2 2 0.2 2 2 2
b8.tif RGB_b432.tif RGB_b432_fused_res15m.tif

Version: 1.0, released 11 June 2013

Details about the options of the tools, including a 3-minute tutorial, can be found here (PDF)

 

 

 

 

 

 

 

(Another brief tutorial contributed by an enthusiastic user)


Tips:

- For a scene with little cloud cover or fewer white objects, these image stretching options (-s 2 0.2 2 0.2 2 0.2 2 2 2) would suffice. If a scene contains a large proportion of white objects (e.g. cloud and snow), try to increase the right-end cut-off values in image histogram stretching for their exclusion. This may be automated along with the estimated percentage of cloud cover over the captured scene, which is indicated in the imagery metadata. For example, if a scene has 20% cloud cover, one may first try to set the right-end cut-off value of 5, about 1/4 of the max % of the cloud cover.
- To improve the brightness of the imagery composite, one may increase non-linear stretching values to 5, 10 or even 20.
- The first pan-sharpening method (1) would perform better for the majority of cases. This choice may be subject to the size of image objects in the scene (i.e. ground features).
- Output files from image stretching and pan-sharpening, if exist, will be over-written automatically.
- The same procedure equally works for any three-band combinations. Through iterations, any number of multispectral bands can be fused and pan-sharpened.
- To avoid typing long file names in DOS console, one can simply drag files to the console.
- One may prepare DOS batch files to efficiently run tools with various image stretching and pan-sharpening options.
- The band combinations and image stretching tools may be used to process the generic 8- or 16-bit imagery from other satellites.
 

Run time:

Spectral Transformer tool set for Landsat-8 imagery is productive. The time to process a full Landsat-8 scene on an average office computer is only about 1 minute (with minimal memory usage). Detailed estimates are as follows:
  • Step 1 (band combination): 1-5 seconds
  • Step 2 (image stretching): 5-20 seconds
  • Step 3 (image pan-sharpening): 15-90 seconds


Other notes:

It is with great privilege that we all have access to such a superior imagery source. We are motivated and have tried our best to develop the tools that can preserve that high quality as much as possible.

We are confident that anyone even with little image processing experience can perform the above analysis. It is our hope that more people can take advantage of and benefit from the excellent Landsat-8 imagery.


(3) Enhanced version with additional features (for Perpetual License)

Two more button clicks: To rapidly make 120 or 336 band combinations using Landsat-8's full spectral bands in a batch mode

  • 120 = 6x5x4, using 6 multispectral bands - Bands 2, 3, 4, 5, 6, 7 (~20GB disk space required for each scene)
  • 336 = 8x7x6, using 8 multispectral bands - Bands 1, 2, 3, 4, 5, 6, 7, 9 (~60GB disk space required for each scene)

A simple yet extremely powerful way of image (visual) analysis: Through rapid, various band combinations, important features/phenomena of interest on Earth land surface can be better highlighted and revealed.

See the invisible!

Innovative applications with numerous band combinations:

- Animated view of 336 band combinations (=8x7x6) of a full Landsat-8 scene (using 8 bands)

- Landsat-8 short-wave infrared (SWIR) bands are capable of detecting wildfire hot spots and lava flows/heat through smokes.

Comparisons: Natural-colour composite (showing full smokes) versus false-colour composite (containing SWIR bands)

Example 1: Happy Camp Complex Fire, California (DATE_ACQUIRED = 2014-09-08).


Example 2: Holuhraun Lava Field, Iceland (DATE_ACQUIRED = 2014-09-06). If your browser (other than Chrome or Firefox) and does not support image slider, full resolution images can be downloaded here or an animation of seeing through smokes is available.


(4) Truly natural-colour, full-scene examples for the "Vivid Earth" series - To view our land more clearly and colourfully!













Full scene overview

Full-scene raw imagery data courtesy of the USGS. All in GeoTIFF format downloaded from the USGS GloVis

 

 

 

 

Comparison and evaluation

Please evaluate the quality of the processed imagery (stretched and pan-sharpened) in terms of:

- spectral consistency
- colour saturation
- colour contrast

- spatial precision

- object detail
- spatial sharpness

- spatial texture

Be the judge on the performance of the tools. Quality Matters!

Full-scene, pan-sharpened result at 15m resolution in GeoTIFF, ready to be used in all GIS and mapping platforms (e.g. ArcGIS, MapInfo) and popular image-editing software (e.g. PhotoShop).

For demonstration, three GIS-ready, compressed formats are included in the zip file:

- JPG (plus *.wld file)
- JPEG2000
- KMZ (Google Earth)


San Francisco
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80440342013106LGN00"

DATE_ACQUIRED=
2013-04-16


stretching option:
-s 2 0.2 2 0.2 2 0.2 5 5 5

Los Angeles
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80410362013133LGN01"

DATE_ACQUIRED = 2013-05-13


stretching option:
-s 2 0.2 2 0.2 2 0.2 5 5 5
New York
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80140322013152LGN00"

DATE_ACQUIRED = 2013-06-01


stretching option:
-s 2 0.2 2 0.2 2 0.2 20 20 20

Washington D.C.
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80150332013111LGN01"

DATE_ACQUIRED = 2013-04-21


stretching option:
-s 2 0.2 2 0.2 2 0.2 10 10 10


The Alps, Europe
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC81960282013115LGN01"

DATE_ACQUIRED = 2013-04-25


stretching option:
-s 2 2 2 2 2 2 20 20 20


Shanghai, China
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC81180382013145LGN00"

DATE_ACQUIRED = 2013-05-25


stretching option:
-s 2 5 2 5 2 5 5 5 5

Sydney, Australia

Newcastle, Australia
(beautiful beach waves + scary river plumes in "3D")


Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80890832013117LGN01"

DATE_ACQUIRED = 2013-04-27


stretching option:
-s 2 0.2 2 0.2 2 0.2 2 2 2

Download all above comparison images Zip file

The size of each processed full scene in a compressed form is only about 5% of that of the raw data in GeoTIFF format.

bands 5/4/3










Equally works for all other band combinations

bands 7/5/3










Download all comparison images Zip file



It only takes about 1 minute to process each full scene on an ordinary computer. Imagine that all daily scenes are processed as above and efficiently delivered as soon as they are available in the following formats (for professionals in remote sensing and GIS, as well as the general public):

  • Conventional GeoTIFF
  • JPEG2000
  • KMZ (Google Earth)
  • Various web mapping tilesets, e.g Google Map Tiles, Bing Maps Tiles, OSM Tiles, TMS Tiles

the potential for the broadest use of the new Landsat-8 imagery is great.


(5) Related applications

Need a global satellite imagery mosaic? High-quality global satellite imagery mosaics (30m-resolution, multiple styles) in ELS2000 V3 series for company-wide, academic campus-wide or Integrator / Internet licenses

Complementary data: e.g. global elevation mosaics and shaded-relief basemaps at 90m and 30m resolutions

Automated, exploratory land cover classifications at 30m and 15m resolutions, at the global scale (ongoing)

Land cover changes

Near real-time / daily monitoring

...


(6) More natural-colour, full-scene examples for the "Vivid Earth" series (added 25 June 2013)














Full scene overview

Full-scene raw imagery data courtesy of the USGS. All in GeoTIFF format downloaded from the USGS GloVis


Comparison and evaluation

Please evaluate the quality of the processed imagery (stretched and pan-sharpened), be the judge on the performance of the tools.

Full-scene, pan-sharpened result:
Two GIS-ready formats are included in the zip file:

- JPEG2000
- KMZ (Google Earth)












Seattle, the U.S.
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80460272013168LGN00"

DATE_ACQUIRED=
2013-06-17

stretching option:
-s 2 20 2 20 2 20 5 5 5











Orlando, the U.S.
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80160402013134LGN03"

DATE_ACQUIRED = 2013-05-14


stretching option:
-s 2 0.2 2 0.2 2 0.2 5 5 5










Rio de Janeiro, Brazil
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC82170762013134LGN02"

DATE_ACQUIRED = 2013-05-14


stretching option:
-s 2 1 2 1 2 1 5 5 5










London, the U.K.
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC82020242013157LGN00"

DATE_ACQUIRED = 2013-06-06


stretching option:
-s 2 10 2 10 2 10 5 5 5











Cairo, Egypt
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC81760392013151LGN00"

DATE_ACQUIRED = 2013-05-31


stretching option:
-s 0.2 0.2 0.2 0.2 0.2 0.2 2 2 2












Beijing, China
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC81230322013132LGN02"

DATE_ACQUIRED = 2013-05-12


stretching option:
-s 2 0.2 2 0.2 2 0.2 2 2 2











Tokyo, Japan
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC81070352013132LGN01"

DATE_ACQUIRED = 2013-05-12


stretching option:
-s 2 5 2 5 2 5 5 5 5











Brisbane, Australia
Zip file   Alternative link

LANDSAT_SCENE_ID = "LC80890792013117LGN01"

DATE_ACQUIRED = 2013-04-27


stretching option:
-s 2 0.2 2 0.2 2 0.2 10 10 10

Download all above comparison images Zip file
The size of each processed full scene in a compressed form is only about 5% of that of the raw data in GeoTIFF format.

 

 

 

Added in 2014

 

 

 

Two formats:

- JPG (plus *.wld file)
- KMZ (Google Earth)

Rome, Italy

Zip file  

LANDSAT_SCENE_ID = "LC81910312013224LGN00"

DATE_ACQUIRED = 2013-08-12

Auto parameterization

Hong Kong and the Pearl River Delta, China

Zip file  

LANDSAT_SCENE_ID = "LC81220442013333LGN00"

DATE_ACQUIRED = 2013-11-29

Auto parameterization

Seoul, South Korea

Zip file  

LANDSAT_SCENE_ID = "LC81160342013259LGN00"

DATE_ACQUIRED = 2013-09-16

Auto parameterization





Disclaimer | Privacy Statement | ©2014 GeoSage