Analyze Broadband Indices

来源: https://vis-webcontent.s3.amazonaws.com/quickguides/Spectral+Analysis/Broadband+Indices/index.html

Lesson content

Lesson 1 of 1

Analyze Broadband Vegetation Indices

Background

Broadband vegetation indices (VIs) measure vegetation health and greenness using broad categories of wavelengths such as red, near-infrared, and green. Of particular interest are red and NIR reflectance. Broadband VIs can be created from multispectral imaging sensors such as Landsat, Sentinel-2, MODIS, PlanetScope, and Quickbird.

Most broadband VIs measure the overall amount and quality of photosynthetic material in vegetation. This, in turn, can indicate the health and vigor of vegetation. ENVI provides 29 different broadband VIs. See the Broadband Greenness(opens in a new tab) topic in ENVI Help for details.

A simple mathematical expression is applied to each pixel in the image to yield a given VI. For example, the Normalized Difference Vegetation Index (NDVI)(opens in a new tab) uses the following equation:

As long as your image has red and near-infrared bands—and their center wavelengths are defined in the metadata—then you can create an NDVI image.

Open a QuickBird Image

1

Select File > Open from the Menu bar. The Open dialog appears. 2. 2

Go to the "data" directory in your ENVI installation path.

Windows: C:\Program Files\NV5\ENVIxx\data (xx is the version number)
Linux: /user/local/NV5/envixx/data
Mac: /Applications/NV5/envixx/data 3. 3

Select the file qb_boulder_msiand click Open. This is a Quickbird multispectral image of Boulder, Colorado. The picture below shows the image at its full extent.

Run the Spectral Indices Tool

1

In the Toolbox, expand the Band Algebra folder and double-click Spectral Indices. The Data Selectiondialog appears. 2. 2

Select qb_boulder_msi and click OK. The Spectral Indices dialog appears. 3. 3

Scroll through the list of VIs to see which VIs are available for this image. Or, click and drag the lower-right corner of the dialog to expand it.

ENVI looks at the wavelength metadata in the associated header file to determine what VIs are available for use. If the image does not have the appropriate band for a given VI, then the VI is not available in this list. For example, the WorldView Improved Vegetative Index is not listed here since that index requires a "NIR 2" band that is only found in WorldView-2 imagery.

4

Select Green Difference Vegetation Index. 2. 5

Hold down the Ctrl key on your keyboard. 3. 6

Select Normalized Difference Vegetation Indexand Simple Ratio. 4. 7

Release the Ctrl key.

8

In the Output Raster field, enter a file name of BroadbandVIs.dat. 2. 9

Click OK. When processing is complete, ENVI creates a multi-band image where each band is a different VI. A color composite is added to the Layer Manager and displayed in the Image window. This image is unnecessary; you will remove it from the view next. 3. 10

In the Layer Manager, right-click on BroadbandVIs.dat and select Remove.

Evaluate the VIs

1

Select File > Data Manager from the Menu bar. The Data Manager appears. 2. 2

Select the Normalized Difference Vegetation Index band, then click the Load Grayscale button. The NDVI band is displayed. This index effectively highlights healthy, green vegetation, but it appears "washed out" with relatively low contrast throughout. However, it is a commonly used VI in remote sensing applications.

3

Select the Green Difference Vegetation Index band, then click the Load Grayscale button. This index is the result of subtracting the green band from the near-infrared (NIR) band. It shows a higher contrast between vegetation and other features, compared to NDVI. For example, the golf course in the top part of the image stands out better against the surrounding landscape when compared to the NDVI image.

4

Select the Simple Ratio band, then click the Load Grayscale button. This index is a ratio between the NIR and red band. It effectively highlights healthy vegetation over a wide range of conditions. The Simple Ratio image shows more variation in greenness, especially in the golf course, compared to the Green Difference Vegetation Index image. In the latter image, the golf course is highly saturated (bright).

5

Compare all three VIs and note the brightness differences in surfaces that contain dormant grass, trees, healthy green grass, etc.

Rather than interpret a grayscale image, you can apply a color table to help visualize the relative health of vegetation.

Apply a Raster Color Slice to NDVI

NDVI images are used in remote sensing to indicate the abundance of vegetation in an area. Their pixel values actually mean something:

•

Values less than 0.1 represent barren rock, sand, snow, and other non-vegetated surfaces. * •

Values of 0.2 to 0.5 indicate sparse vegetation such as shrubs, grasslands, scenescing crops, etc. * •

Values of 0.6 to 0.9 represent thick, dense vegetation typically found in tropical forests or crops at their peak growth.

Source: Brown, Jesslyn. "NDVI, the Foundation for Remote Sensing Phenology." Updated November 27, 2018. https://www.usgs.gov/special-topics/remote-sensing-phenology/science/ndvi-foundation-remote-sensing-phenology(opens in a new tab).

Applying a color table (called a raster color slice in ENVI) to an NDVI image can help you visualize these data ranges. For this exercise, you will restore a color slice that was already created for you.

1

In the Layer Manager, uncheck the Simple Ratio and Green Difference Vegetation Index layers to hide them. 2. 2

Right-click on the Normalized Difference Vegetation Index layer and select New Raster Color Slice. The Data Selection dialog appears. 3. 3

Select the Normalized Difference Vegetation Index band and click OK. The Edit Raster Color Slices dialog appears. A default color slice is applied to the NDVI image. We do not want this color scheme, so you will remove it in the next step. 4. 4

In the Edit Raster Color Slices dialog, click the Clear Color Slices button.

5

In the same toolbar, click the Restore Color Slices from File button. A file selection dialog appears. 2. 6

Go to the "hyperspectral" directory in your ENVI installation path.

Windows: C:\Program Files\NV5\ENVIxx\data\hyperspectral (xx is the version number)
Linux: /user/local/NV5/envixx/data/hyperspectral
Mac: /Applications/NV5/envixx/data/hyperspectral 3. 7

Select the file RENDVIColorSlice.dsr and click Open. The raster color slice is displayed over the NDVI image.

This color slice was originally designed for Red Edge NDVI images, which are derived from hyperspectral images. However, its data values apply to NDVI images as well.

The color slice shows that most of the vegetation in this scene falls within the 0.2 to 0.5 range. Indeed, the QuickBird image from which NDVI was derived was acquired in the winter months. Nearly all of the vegetation is either dormant or sparse. Only a few pixels are above 0.5. These correspond to evergreen trees.

8

This concludes the exercise.

In summary, broadband VIs use broad wavelength ranges—such as red, blue, green, NIR, and SWIR—to indicate the presence and relative abundance of healthy green vegetation. Each VI was originally designed for a unique purpose, so they do not all yield the same results but are nonetheless effective at characterizing vegetation as a whole in a study area.

Additional Resources

•

Analyze Narrowband Vegetation Indices quick guide

Your input is important to us, please take a few moments to fill out ourQuick Guide Feedback(opens in a new tab)form.