Merge pull request #321 from azlinszkysinergise/main

Edited the Sentinel-2 True Colour custom script to also include current EOB Default

sentinel-2/true_color/README.md

@@ -5,13 +5,18 @@ grand_parent: Sentinel
 layout: script
 permalink: /sentinel-2/true_color/
 nav_exclude: true
+scripts:
+- - Visualization
+  - script.js
+- - EO Browser
+  - eob.js
 examples:
-- zoom: '10'
-  lat: '42.01869'
-  lng: '12.58484'
+- zoom: '12'
+  lat: '34.23664'
+  lng: '127.20074'
   datasetId: S2L2A
-  fromTime: '2023-10-12T00:00:00.000Z'
-  toTime: '2023-10-12T23:59:59.999Z'
+  fromTime: '2024-05-16T00:00:00.000Z'
+  toTime: '2024-05-16T23:59:59.999Z'
   platform:
   - CDSE
   - EOB
@@ -23,7 +28,18 @@ examples:
 
 ## General description
 
-The true color product maps Sentinel-2 band values B04, B03, and B02 which roughly correspond to red, green, and blue part of the spectrum, respectively, to R, G, and B components.
+The true color product maps Sentinel-2 band values B04, B03, and B02 which roughly correspond to red, green, and blue part of the spectrum, respectively, to R, G, and B components of the image shown.
+
+But what color is true color? Various additional processing steps can be applied to scale the image values in a range that is well perceived and pleasing to the eye [(Sovdat et al, 2019)]((https://www.sciencedirect.com/science/article/pii/S0034425719300422)).
+
+- For the default script here, we apply simple gain correction, multiplying the pixel values of each image band with a single constant value. This is the `gain` parameter which can be adjusted in the script window.
+
+- For [Copernicus Browser](https://browser.dataspace.copernicus.eu/), we chose to apply default a true color script which adds [Highlight Compression](https://docs.sentinel-hub.com/api/latest/evalscript/functions/#highlightcompressvisualizer) and also saturation and contrast adjustment. In this context, saturation is the brightness of colours in the image and contrast is the difference between the bright and dark areas.
+The current True Color algorithm for Copernicus Browser can be accessed as [Sentinel-2 L2A optimized True Color](https://custom-scripts.sentinel-hub.com/sentinel-2/l2a_optimized/) and is also compatible with EO Browser.
+
+- The default True Color script in [EO Browser](https://apps.sentinel-hub.com/eo-browser/) is more simple, with Highlight Compression between specific `min` and `max` values. The `min` and `max` values are user-adjustable, and of course, this script also works in Copernicus Browser
+
+The bands to use for True Color are
 
 For Sentinel-2: **BO4, B03, B02**
 
@@ -37,10 +53,20 @@ For [MODIS](https://custom-scripts.sentinel-hub.com/modis/true-color/): **B01, B
 
 ## Description of representative images
 
-True color visualization of Rome, on 8.10.2017.
+True color visualization of islands in the Korea Strait
+- With basic True Color Product (this script)
+
+![True color visualization of islands in the Korea Strait, on 16.05.2024.](fig/basic_true_color.jpg)
+
+- With EO Browser default True Color
+
+![True color visualization of islands in the Korea Strait, on 16.05.2024.](fig/eob_default_true_color.jpg)
 
-![True color visualization of Rome, on 8.10.2017.](fig/fig1.png)
+- With Copernicus Browser default True Color [Sentinel-2 L2A optimized True Color](https://custom-scripts.sentinel-hub.com/sentinel-2/l2a_optimized/)
 
+![True color visualization of islands in the Korea Strait, on 16.05.2024.](fig/cdse_default_true_color.jpg)
 
 ## References
+ - Sovdat, Blaž, Miha Kadunc, Matej Batič, and Grega Milčinski. "Natural color representation of Sentinel-2 data." Remote sensing of environment 225 (2019): 392-402. [Full Text](https://www.sciencedirect.com/science/article/pii/S0034425719300422)
+
  - Wikipedia, [True color](https://en.wikipedia.org/wiki/False_color#True_color). Accessed October 10th 2017.

sentinel-2/true_color/eob.js

@@ -0,0 +1,17 @@
+//VERSION=3
+let minVal = 0.0;
+let maxVal = 0.4;
+
+let viz = new HighlightCompressVisualizer(minVal, maxVal);
+
+function setup() {
+   return {
+    input: ["B04", "B03", "B02", "dataMask"],
+    output: { bands: 4 }
+  };
+}
+
+function evaluatePixel(samples) {
+    let val = [samples.B04, samples.B03, samples.B02, samples.dataMask];
+    return viz.processList(val);
+}