Hi Nigel,
I have searched for a tool to warp an image once, in a kind of edit mode, much like the TPSShapeTransform demo project, and then apply that transformation to images of the same size. From the IEVision documentation I understand that there is a possibility to predefine the two kinds of points (as explained in TIEVisionImage.thinPlateSplineShapeTransform of the help doc).
In that case one has to define two sets of IEVisionLib().createVectorPoint2f() and apply these to the GetIEVisionImage().thinPlateSplineShapeTransform where it then generates a new TIEVisionImage, which in turn can be assigned to an ImageEnView.
That is an attractive solution for me. However, I don't want to define these two sets of points, but derive them from my visual warp work in the GUI.
I get the mechanism, but I actually want to extract these points from an GUI interaction, where the warp grid is created; then extracted these two sets of gridpoints, save them, and re-use them on another image (of same dimensions). Basically, if I use 12 points, I want to get these 12 points, both the starting (X,Y) and the destination (X,Y), from the GUI.
My question: how can I get these two sets of points from the GUI after I have manually prepared the warp transformation?
In case this helps to explain my question better: this is provided in the help doc:
img := ImageEnView1.IEBitmap.GetIEVisionImage().thinPlateSplineShapeTransform( srcPoints, dstPoints, ievBORDER_REPLICATE, IEVisionScalar( 0,0,0 )); where the two sets (srcPoints and dstPoints) are predefined. But I need them derived.
Thanks for your help,
Sybren |
Search
TPSShapeTransform: extract the warping grid points
Posted - May 29 2026 : 16:23:27
