Image Statistics

The Overview and Try It sections provide a basic understanding of the process. Subsequent sections have more in–depth information and can be used as a reference. A step by step tutorial lesson is included.

Overview

The Image Statistics process computes raster cell statistics from one or more selected images. You can choose to compute standard statistical values such as mean, median, and standard deviation, and also compute a number of Gray Level Co-occurrence Matrix statistics.

Try It

Prerequisite Skills: Getting Started and Displaying Geospatial Data.

Sample Data: stanton_landsat8.rvc in stanton.zip

Tutorial Lesson: The exercise below provides an introduction.

input

The Image Statistics process.

Exercise 1 – Run the Image Statistics process

To run the process: load the rasters, select the statistics you want, and run.

  1. select Image > Statisics > Statistics by Band from the TNTmips menu bar
  2. click the Select button and add all of the rasters in the stanton_landsat8.rvc file
  3. select the statistics you want
  4. click the Run icon and when prompted choose .csv or .txt for the New File Type and name the output file
  5. look at the output file in a text editor or spreadsheet program

input

Resulting .csv file.

Input

Raster objects

The process takes one or more grayscale, hyperspectral, or color composite images as input.

Output

Statistics can be output to a .txt or .csv file. The process outputs a separate row of statistics for each color component or image band.

Interface

The Image Statistics window is used to run the process.

Top Toolbar

Run – run the process after selecting the desired statistics.

Queue Job / Save Job – standard Queue Job and Save Job icon buttons allow you to execute the process using the TNTgis Job Processing system. See Job Processing for documentation.

Preferences – turn off the Allow spaces in field names option to replace spaces with the underscore character.

Help – process documentation.

Images pane

This section lets you select the input and shows it in a raster input list.

Select – choose one or more images from which to compute the statistics. You can choose grayscale, hyperspectral, or color composite images.

Remove / Remove All – remove selected or all images from Images list.

Statistics pane

The Statistics box lists the measures available for the selected element type with a checkbox allowing you to select or deselect each. The options include: Cell Count and standard statistical measures: Minimum, Maximum, Range, Mean, Median, Mode, Standard Deviation, and Variance.

Relative Std Deviation % – Is a dimensionless value computed by dividing the standard deviation by the mean and multiplying by 100 to express the value as percentage. This value is generally only meaningful when all cell values are positive.

Gray Level Co-occurrence Matrix (GLCM) – A number of Gray Level Co-occurrence Matrix (GLCM) texture measures are available. These measures first quantize the range of image cell values into bins using the value in the Number of GLCM bins field (range 16 to 64). A co-occurrence matrix is computed to record how often specific pairs of bin values occur in immediately neighboring cell positions (horizontally and vertically).

Each GLCM option derives a different statistical texture measure from that matrix. For details of the individual GLCM measures, see the Technical Guide entitled Raster and Image Processing: Gray Level Co-occurrence Matrix Filters.