Correcting Row and Column Shifts in a Calibration Matrix for a Projection System

A method for correcting a projection from a plurality of digital projectors, comprising:

• receiving as an input a matrix which maps each pixel to be projected to a captured pixel that includes a row value and a column value;
• detecting a shift with respect to one of row or column values in the matrix, wherein the shift is measured by an overlap length;
• determining a gradient length from the overlap length;
• generating a gradient of pixel weightings for each gradient length;
• merging any gradients that intersect, wherein the merging results in the creation of a list of that includes common points and diagonal points;
• and
• updating the captured values in the matrix with new values determined by the list.

A non-transitory computer-readable storage medium with an executable program stored thereon, wherein the program instructs a microprocessor to perform the following steps:

• receiving as an input a matrix which maps each pixel to be projected to a captured pixel that includes a row value and a column value;
• detecting a shift with respect to row or column values in the matrix, wherein the shift is measured by an overlap length;
• determining a gradient length from the overlap length;
• generating a gradient of pixel weightings for each gradient length;
• merging any gradients that intersect, wherein the merge results in the creation of a list of that includes common points and diagonal points;
• and
• updating the captured values in the matrix with new values determined by the list.