Point Statistics Raster

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 Point Statistics Raster process creates a raster with cell values reflecting the point count or point statistics within an area. Thus a point statistics raster provides a reflection of the distribution of point features. Point data that covers a wide area may appear nearly continuous and uniform. However, if these points represent some naturally occurring feature or phenomenon, rather than a regular sampling, there will be areas where the points are denser than others. Statistics are accumulated from a collection of points by determining which points fall within each raster cell or a specified radius of the cell center. The points may simply be counted to produce the cell value or density (points per unit area), or attribute fields can be specified to provide values to use in computing various statistics.

Try It

Prerequisite Skills: Getting Started and Displaying Geospatial Data.

Sample Data: the nebraska_towns.rvc file in nebraska.zip and ARRA-AL_TennesseeRiver_2010_000002_ground.las in alabama.zip

Tutorial Lesson: The exercises below provide an introduction.

Tutorial Lessons:

These lessons provide a quick introduction.

  1. Point count within radius
  2. Attribute statistics for points within radius
  3. Attribute statistics for LiDAR points in cell

input

The Point Statistics Raster process.

Exercise 1 – Point count within radius

To run the process: load the input points, set output raster grid (image size), set Accumulation method and run for Total Points or other output option.

In this exercise, points are counted within the specified radius and used for the raster cell values. The Accumulate by Radius option can be useful when working with sparse points.

  1. select Image > Spatial Analysis > Point Statistics Raster from the TNTmips menu bar
  2. click the Points button and add the nebraska_towns.rvc / towns vector

    3. Leave the Image Grid Alignment set to User Defined and Extents, Selection: Input.

  3. set Image Size, Rows: 1000 (Columns and Cell Size settings will adjust automatically)
  4. in the Output panel, set Accumulate to Radius: 15 km.

    5. Leave the Color Spread: Manual and Path: Automatic.

  5. turn on Statistics from Point Counts: Total Points and Set Zero as Null
    6. turn off Density toggle
  6. click Run
  7. look at the new raster added to the View window

input

The input is a vector object with point elements.

input

Resulting statistics raster after running.

input

Output cell values indicate how many points are within 15 km. The output raster values tend to show a radius–like pattern around the points. In this example, the radius appears in an elliptical shape instead of a circle due to using the Geographic coordinate reference system.

(continued from previous exercise)

Exercise 2 – Attribute statistics for points within radius

Instead of counting the points, compute statistics from attributes of points within specified radius.

  1. in the Input section, limit the points using the By Query option
  2. in the Script Editor window that opens enter the following: 'builtupp_usa.pop > 0' (without quotes)

    3. Note, you can use the Insert Field option to select the table and field. The 'pop' field contains the town populations.

  3. turn off all toggles in the Statistics from Point Count section
  4. in the Statistics from Attribute Fields section use Add Field and select the table and field: builtupp_usa.pop
  5. turn on the Sum toggle
  6. click Run
  7. look at the results in the View window

input

Output raster created using town population values.

input

Accumulated population values for towns (points) within 15 km radius of cell center. Cells within 15 km of multiple towns get the total population of all those towns.

Exercise 3 – Attribute statistics for LiDAR points in cell

Instead of gathering points within a radius, the Accumulate: By Cell option looks at points within the raster cell. The Accumulate by Cell option is useful when working with dense points.

    The data provided for this exercise requires a TNTmips Pro license. If you don't have a professional license, use the nebraska_towns.rvc / towns vector with modified settings: set the cell size large enough to contain multiple points and select a numeric attribute such as population (builtupp_usa.pop).

  1. click the Points button and add the alabama\ARRA-AL_TennesseeRiver_2010_000002_ground.las file
  2. leave the Image Grid Alignment set to User Defined and the Extents, Selection option to Input
  3. set Cell Size: 10 (Column and Image Size settings will adjust automatically)
  4. in the Output section set Accumulate: Cell and leave Color Spread set to Manual with Path: Automatic
  5. turn off all toggles in the Statistics from Point Count section
  6. in the Statistics from Attribute Fields section use Add Field and select the table and field: ARRA-AL_TennesseeRiver_2010_000002_ground.Z
  7. turn on the Mean toggle
  8. click Run
  9. look at the results in the View window
  10. change Accumulate to a Radius of 15 meters, the Cell Size to 1x1 meter, run the process again, and compare results

input

LiDAR input.

input

Results raster from mean of point attributes (Z). Values were accumulated for points within each output cell. The output raster was given a large Cell Size of 10x10 meters.

input

Results raster from mean of point attributes (Z). Values were accumulated for points within a 15 meter radius around the cell centers. The output raster has a Cell Size of 1x1 meter.

Input

Geometric Points

The process takes point elements as input (from a vector, CAD, shape object, shapefile, or .las LiDAR file). You have the option to include all elements or limit them using a database query. The points within the output raster cell or specified radius can be counted. In addition, values in the point database can also be used to figure cell values.

Results & Output

The process outputs different types of statistics rasters. The data type will be determined on the range of cell values. The process includes the option of setting a color palette that is used when displaying the result.

There are three general types of statistics rasters you can make: Total Points (point count), Density, or from Attribute Fields. Set Accumulate to Cell to include each point in only the cell that contains it. In this case, every point will be counted exactly once, assuming no points fall outside the output raster extents. When Accumulate is set to Radius all points within the designated radius of the cell center are assigned to that cell. Generally this will result in points being counted for multiple cells. It is recommended that the radius be at least the distance from the cell center to the corner to avoid excluding points from being counted.

Interface

The Point Statistics Raster window is used to run the process.

input

The Point Statistics Raster process.

Input pane

Points – Select a geometric object or shape file with point elements. Specify or limit the points of interest using the All and By Query options.

Image Grid Alignment, Extents, Cell Size, and Image Size panes

Specify the output raster's grid, extents, cell size, and image size using these four panes. For more information, see the Raster Grid Controls section.

Output pane

Accumulate – Specify the method used to count points, either by Cell or Radius. Then if using the Radius option, enter the size of the radius.

Color Spread – Menu to select method on how to create the output raster color palette: None, Manual, Grayscale (linear), or Choose. The Manual option gives you two color boxes to set the start and end colors, which is used with the Path option to create the colors: Automatic, RGB, HIS Clockwise, HIS Counter–Clockwise, HBS Clockwise, and HBS Counter–Clockwise. The Choose option opens the Select Palette window. You also have the None and Grayscale options.

Note, if you choose the Manual method, the colors are based on the cell value range and thus may be different from one input to another. However, if you choose a color palette, the colors are set based on the color order in the palette and will remain the same for the same values over different datasets.

Statistics from Point Counts pane

Total Points – Point counted based on Accumulate option: the number of points within the output Cell or specified Radius.

Density – Points per unit area.

Note that if Accumulate is by Cell and Density is Points per Cell, the Total Points and Density results will be equal. For areas where no points occur, the Set Zero as Null toggle allows these cells to be "null" to indicate no data, and typically display as transparent. The actual output cell type will be determined automatically based on the maximum count occurring in the resulting raster.

Statistics from Attribute Fields pane

Add Field... – Select one or more field(s) to get numeric point attributes. Then, any or all of the available statistics can be computed: Sum, Mean, Minimum, Maximum, Range, Variance, and Standard Deviation.

Note, the colors are always spread over the range of output values, regardless of whether a manual or predefined palette is used.