List of Combo Overlay examples

From Tygron Support wiki
Jump to navigation Jump to search


Attribute

Solid Buildings

The attribute SOLID generally has a value of 0 or 1. It can therefore be used directly in a Combo Overlay Formula.

  • Set Combo Maplink to BUILDINGS
  • Set Key A to SOLID
  • Set Formula to:
    @A

Residential Buildings

The unit size category values can be used to identify and rasterize residences, in combination with a GT (Greater than) function with multiple arguments. To setup:

  • Set Combo Rasterization to "Attribute:Specific Layer" (SINGLE_LAYER)
  • Set Combo Maplink to Buildings
  • Set Key A to SOCIAL_UNIT_SIZE_M2
  • Set Key B to NORMAL_UNIT_SIZE_M2
  • Set Key C to LUXE_UNIT_SIZE_M2
  • Set Key D to STUDENT_UNIT_SIZE_M2
  • Set Key E to SENIOR_UNIT_SIZE_M2
  • Set Formula to:
    GT(@A,@B,@C,@D,@E)

Categorized Residential Buildings

The unit size category values can be used to identify and rasterize residence types, in combination with a IF and GT (Greater than) functions with multiple arguments. We will categorize the residential buildings as followed: SOCIAL (1) , NORMAL (2), LUXE (3), SENIOR (4), STUDENT (5) and everything else 0.

To setup:

  • Set Combo Rasterization to "Attribute:Specific Layer" (SINGLE_LAYER)
  • Set Combo Maplink to Buildings
  • Set Key A to SOCIAL_UNIT_SIZE_M2
  • Set Key B to NORMAL_UNIT_SIZE_M2
  • Set Key C to LUXE_UNIT_SIZE_M2
  • Set Key D to SENIOR_UNIT_SIZE_M2
  • Set Key E to STUDENT_UNIT_SIZE_M2
  • Set Formula to:
    IF(GT(@A,0),1,IF(GT(@B,0),2,IF(GT(@C,0),3, IF(GT(@D,0), 4, IF(GT(@E,0),5,0))))

Rasterization

Minimum infiltration among layers

An example on how the combo overlay can obtain rasterized data from multiple data layers (MapLinks); Terrains, Buildings, Neighborhoods, Areas and Zones.

Timeframes

Minimum value among timeframes

An example on how the Combo Overlay can calculate the minimum value among timeframes of a Prequel Grid Overlay

Maximum value among timeframes

An example on how the Combo Overlay can calculate the maximum value among timeframes of a Prequel Grid Overlay

Average value among timeframes

An example on how the Combo Overlay can calculate the average value among timeframes of a Prequel Grid Overlay, for example a Shadow Overlay:

Mask for each timeframe

An example on how the Combo Overlay can calculate, for each timeframe of an input overlay, whether a grid value is above a specific value. In this example a Heat Stress Overlay with a minimum value of 45 degrees.

Filter each timeframe

An example on how the Combo Overlay can calculate, for each timeframe of an input overlay, only include a grid value when it is above a specific value, and make it NO_DATA otherwise. In this example a Heat Stress Overlay with a minimum value of 45 degrees.

  • Set A prequel (Combo Overlay) to a Grid Overlay, in this example a Heat Stress Overlay.
  • Set Formula to:
    IF(GT(AT, 45), AT, NO_DATA)

Iteration

For initializing a Combo Overlay in a iteration setup, the ITERATION parameter can be used to obtain the active Iteration number. In this example we initialize the value to 0.

Inference Analysis

You can validate the effectiveness of a trained RCNN using an Inference Overlay by comparing its recognized features with manually drawn features (using Areas with a particular Attribute) and perform Intersection over Union calculations. It can also be useful to visualize regions where features were recognized but not drawn manually, and vice versa. You can use the following setup to do these analyses. In this example we will use a RCNN for recognizing foliage, and have drawn Areas manually and assigned the attribute FOLIAGE to them. Base setup:

Union function:

Colors

Grid input to colors

An example on how a combo overlay can return a complex colorization using a grid input without using a legend. This example expects the value of an input Grid Overlay to be:

  • Negative or NO_DATA in case it has no direction
  • A value between 0 and 360 indicating a direction.

Configure the Combo Overlay as followed:

  • Set Grid A to an input overlay representing a direction in degrees, such as Surface avg direction result type (Water Overlay)
  • Set Formula to:
    IF(OR(LT(AT, 0), EQ(NO_DATA, AT)), NO_DATA, IF(LT(AT, 15), COLOR(255, 0), IF(LT(AT, 45), COLOR(255, 125, 0), IF(LT(AT, 75), COLOR(255, 255, 0), IF(LT(AT, 105), COLOR(125, 255, 0), IF(LT(AT, 135), COLOR(0, 255, 0), IF(LT(AT, 165), COLOR(0, 255, 125), IF(LT(AT, 195), COLOR(0, 255, 255), IF(LT(AT, 225), COLOR(0, 125, 255), IF(LT(AT, 255), COLOR(0, 0, 255), IF(LT(AT, 285), COLOR(125, 0, 255), IF(LT(AT, 315), COLOR(225, 0, 225), IF(LT(AT, 345), COLOR(225, 0, 125), COLOR(255, 0, 0))))))))))))))

Saturate Satellite Imagery

The satellite imagery colors can be saturated using a saturation approximation function. In this example the factor is set to 0.9. Attributes are stored in the Combo Overlay and used in the formula with the prefix GLOBAL_

  • Set Grid A to a Satellite Overlay
  • Add Global named R_FACTOR with value 0.2989
  • Add Global named G_FACTOR with value 0.587
  • Add Global named B_FACTOR with value 0.114
  • Add Global named SATURATION with value 0.9
  • Set Formula to:
    COLOR(
     ADD(MUL(-1, GLOBAL_SATURATION, 
      ADD(MUL(GLOBAL_R_FACTOR, RED(A)), MUL(GLOBAL_G_FACTOR, GREEN(A)), MUL(GLOBAL_B_FACTOR, BLUE(A)))), MUL(ADD(1, GLOBAL_SATURATION), RED(A))), 
     ADD(MUL(-1, GLOBAL_SATURATION, 
      ADD(MUL(GLOBAL_R_FACTOR, RED(A)), MUL(GLOBAL_G_FACTOR, GREEN(A)), MUL(GLOBAL_B_FACTOR, BLUE(A)))), MUL(ADD(1, GLOBAL_SATURATION), GREEN(A))), 
     ADD(MUL(-1, GLOBAL_SATURATION, 
      ADD(MUL(GLOBAL_R_FACTOR, RED(A)), MUL(GLOBAL_G_FACTOR, GREEN(A)), MUL(GLOBAL_B_FACTOR, BLUE(A)))), MUL(ADD(1, GLOBAL_SATURATION), BLUE(A)))
    )

The factors for red, green and blue are based on YCbRc ITU-R BT.601 conversion[1].

Color Filtering

For feature recognition using Inference Overlays with RCNNs, it might be useful to narrow the space for recognition by making parts of the satellite imagery black where it is not needed. This can be used for both for the initial training and test data, as well as direct input for Inference Overlays. In this example, we are interested in gardens of residential areas that are marked as private yards.


See also

References

  1. YCbRc ITU-R BT.601 conversion ∙ Found at: https://en.wikipedia.org/wiki/YCbCr#ITU-R_BT.601_conversion ∙ (last visited: 2026-06-30)