I am trying to use this function and find the documentation a little confusing. Document does not state the size of the output image. When I looked in the example sources, I see a usage such as :
VLIB_xyGradientsAndMagnitude(IMAGE1, GRADX+64+1, GRADY+64+1, GRADMAG+64+1, 64, 64);
So this is pointing the output pointers to element [1][1] essentially.
What is not stated is how the ends of the rows or the last line are treated. I tried filling my output buffer with 0xDEAD before calling and the beginning of the output is unchanged, with first non 0xDEAD coming at width+1 elements in . but I do not see any gaps at row ends. I want to be sure there is no skewing happening. If call the routine with the full height, the output is modified past the end of my buffer. If I call it with height-2, as I noticed in the Canny example, I get exactly one line of un touched buffer. Which I believe means I have no skewing going.
Can someone confirm that I have this correctly?
Is there any better documentation then SPRUG00C? its very inconsistent in depth of information among the various methods.
Anybody out there of any landmines in other methods....
thanks
Dave