OEPNTTHIN

Description

OEPNTTHIN thins the points in a CATALYST Professional OrthoEngine project file by automatically removing redundant tie points (TP), ground control points (GCP), or both.

Parameters

oepntthin(oeproji, oeprojo, ptype, numgrid, ppcell, pointopt, residual)

Name	Type	Caption	Length	Value range
OEPROJI *	str	Input OrthoEngine project file	1 -
OEPROJO *	str	Output OrthoEngine project file	1 -
PTYPE *	str	Point Type	1 -	BOTH \| GCP \| TP Default: BOTH
NUMGRID *	List[int]	Grid size for thinning	1 -	3 - 1000 Default: 3,5
PPCELL *	List[int]	Points per cell	1 -	1 - 1000 Default: 1,2
POINTOPT *	str	Point option	3 -	INACT \| DEL Default: INACT
RESIDUAL *	str	Residual calculation	1 -	YES \| NO Default: NO

* Required parameter

Parameter descriptions

OEPROJI

The name of the input OrthoEngine project file that contains the points to refine.

OEPROJO

The name of the output OrthoEngine project file.

PTYPE

The type of points in the OrthoEngine project file to thin.

The supported values are as follows:

GCP: Apply thinning only to GCPs
TP: Apply thinning only to TPs
BOTH: Apply thinning to GCPs and TPs

Note: GCP thinning is not applicable to airborne digital sensor (ADS) project files. With ADS project files, thinning will be applied only to TPs, regardless of the setting for PTYPE.

NUMGRID

The number of grid cells to use when dividing the image extents for the thinning process.

The parameter is defined as follows:

			NUMGRID=PARAM1, PARAM2

You can specify one or two values for this parameter.

If you specify a single value, it will be used for thinning of both GCPs and TPs (based on what you specify for PTYPE). If you specify two values, the first applies to thinning GCPs and the second applies to thinning TPs, regardless of what is specified for PTYPE.

PARAM1=3: Represents "the square image will be divided into three-by-three grid for GCP thinning".

Note: PARAM1 = 3,4,5,...; must be 3 or greater.
PARAM2=5: Represents the image is divided into five-by-five grids for TP thinning.

Note: PARAM2 = 3,4,5....; must be 3 or greater.

The value entered will be used for both the x and y direction if the imagery is square. If the image is not square, the number entered will be used on the short-side of the image and the long side will be calculated automatically.

For example, if you specify 3,5, the image will be divided into a three-by-three grid for GCP thinning and a five-by-five grid for TP thinning. If you enter a single value of 5, the image will be divided into a five-by-five grid for thinning both GCPs and TPs.

Note: With an ADS project, the values for NUMGRID are ignored, because OEPNTTHIN automatically divides the imagery into segments, based on the overall length of the strips.

Default values are 3,5

PPCELL

The number of points to retain after thinning each grid cell.

The parameter is defined as follows:

		PPCELL=PARAM1, PARAM2

You can specify one or two values for this parameter.

If you specify a single value, it will be used for thinning of both GCPs and TPs (based on what you specify for PTYPE). If you specify two values, the first applies to thinning GCPs and the second applies to thinning TPs, regardless of what is specified for PTYPE.

PARAM1=1: Represents "each cell retains one GCP."

Note: PARAM1 = 1,2,3,4,5,...; must be 1 or greater.
PARAM2=2: Represents "each cell retains two TP points".

Note that PARAM2 = 1,2,3,....; must be 1 or greater.

Default values are 1,2.

Note: With ADS data, the TP value of PPCELL defines the number of points to retain in each segment. Generally, you will use a much higher value than with non-ADS projects.

POINTOPT

Whether to change redundant points to inactive points or remove them. Supported values are:

DEL: removes the points from the project.
INACT: retains the points in the project, but renders them inactive. Inactive points do not factor in the model calculations.

INACT is the default.

RESIDUAL

Whether to calculate residual values during thinning.

Supported values are:

YES: calculate residual values during thinning. OEPNTTHIN calculates the residual value for each point in the project using the input math-model information. This information is then used as a decision factor when comparing points within a grid cell.
NO: do not calculate residual values during thinning. OEPNTTHIN uses the number of overlaps as the main methodology for selecting which points to retain in the grid cells.

Return Value

Returns: Execution status

Type: PCI_INT

The return value is 0. This function returns only if it executes successfully; otherwise, it throws an exception.

Details

OEPNTTHIN removes redundant TPs and GCPs from an OrthoEngine project.

Automatic TP or GCP collection sometimes provides too many points in each image, and the point distribution may vary. OEPNTTHIN can reduce the number of points while maintaining a more consistent distribution of points across the image. This is achieved by dividing the image into a configurable number of grid cells and selecting points to retain in each grid cell.

OEPNTTHIN reduces the number of points as follows: with each grid cell, the points with the lowest connectivity (smaller number of rays) are removed, leaving the desired number of points in the cell. In this sequence, the points with the best connectivity will remain. If the you select to use residuals for thinning, the residual will also be considered. Points with the lowest connectivity and highest residuals are prioritized for removal.

OEPNTTHIN can be set up to designate the removed points as inactive, or remove them. You can manually examine the points designated inactive in OrthoEngine to decide if they should be removed.

After running OEPNTTHIN, an even distribution of points and proper number of points can be achieved. The model computation time and the error-detection effort should be reduced significantly.

Example

The following example removes both redundant GCPs and TPs. With GCP thinning, each image will be divided into three-by-three grids, with one point in each grid cell. With TP thinning, the image will be divided into five-by-five grids, with two points in each grid cell. Redundant points will be removed. Redundant points will be identified by residual calculations.

from pci.oepntthin import oepntthin

oeproji	=	"airphoto_model.prj"
oeprojo	=	"a.prj"
ptype   =       "BOTH"
numgrid =       [3,5]
ppcell  =       [1,2]
pointopt=       "INACT"
residual=       "YES"

oepntthin (oeproji, oeprojo, ptype, numgrid, ppcell, pointopt, residual)

Environments	PYTHON :: EASI :: MODELER
Quick links	Description :: Parameters :: Parameter descriptions :: Return Value :: Details :: Example