TPREFN

MODE=1 represents the absolute number of TPs to reject.

For example, to give twice the weight to the x-residual as the y-residual, enter a rejection weighting of 2. By default, the residual in x and y will have the same weight. Therefore, if you have a point with an x-residual of 0.4 and a y-residual of 0.5, the point will be given an x-residual of 0.8 and a y-residual of 0.5 for the rejection.

To, for example, apply the Absolute Number rejection method (mode 1), reject 10 TPs, and give the x-residual half the weight as that of the y-residual, enter 1, 10, 0.5.

MODE=2 represents the percentage number of points to reject, starting with the one with the highest residual.

For example, to give twice the weight to the x-residual as the y-residual, enter a rejection weighting of 2. By default, the residual in x and y will have the same weight. Therefore, if you have a point with an x-residual of 0.4 and a y-residual of 0.5, the point will be given an x-residual of 0.8 and a y-residual of 0.5 for the rejection.

To, for example, apply the Percentage Number rejection method (mode 2), reject five percent of GCPs, and give the x-residual twice the weight as that of the y-residual, enter 2, 5 ,2.

MODE=3 represents the sigma standard deviation of points to reject. This mode determines the points with the largest x and y residuals at each iteration, and computes standard deviations of x and y residuals of all remaining points. A TP is rejected if it has the largest root-mean-square (RMS) residual, and its x-residual or y-residual value is greater than PARAM1 or PARAM2 multiplied by the x-sigma or y-sigma value, respectively.

For example, to reject points with an x-residual value higher than the x-sigma multiplied by two, and to reject points with a y-residual higher than the y-sigma of the remaining points, enter 2 for PARAM1 and 1 for PARAM2.

MODE=4 represents the absolute distance of the points to reject in the residuals. At each iteration, the point with the largest x or y residual is rejected, if the respective residual is greater than PARAM1 or PARAM2, respectively.

Rejection starts from the point with the largest x or y residual distance.

For example, to apply the Absolute Distance rejection method (mode 4), reject pixels with an x-residual value greater than two pixels, and reject pixels with a y-residual value greater than two pixels, enter 4, 2, 2.

For this mode TPREFN will check deleted points after refinement and repeat again for those points with errors less than 2 times the specified rejection pixels. This is in case good TPs are deleted initially due to other large TP errors.

MODE=5 represents the RMS error of points to reject. Rejection starts from the point with the largest residual error for any point, then recomputes the math model and RMS error. If the RMS error is still above the specified thresholds, the point with the next highest residual is removed and so on, until the x-RMS and y-RMS errors are equal to or less than PARAM1 pixels or PARAM2 pixels.

For example, to apply the RMS Error rejection method (mode 5) and reject points with the highest residuals until the x-RMS and y-RMS values are both less than two pixels, enter 5, 2, 2.

MAXPT is the maximum number of TPs to remove in each iteration before recomputing the model. A positive value specifies explicitly the number of TPs that can be removed. A value of 0 removes the number of TPs relative to the number of images in the project. With a negative value, its absolute value specifies the percentage of TPs that can be removed. Regardless of the value you specify, at least one point per iteration is always allowed to be removed.

MINPT is the minimum number of points per image to preserve. That is, this ensures that on completion of processing each image has at least the minimum number of TPs remaining that you specify.

You can specify MINPT as an optional second value with mode 6.

REMOVE_BLUNDER specifies whether to use use RANSAC or initial model to remove large blunders before refinement. 0: NO, 1: RANSAC, >1: INITIAL_MODEL. If REMOVE_BLUNDER is INITIAL_MODEL, it should be a blunder threshold in which large blunders are to be removed using initial model.