I am writing a program using Image Resample VI. Since it comes with Vision Development Module (I only have a trial license for it, however full license for Vision acquisition software), I tried using the Interpolate 2D.vi directly manipulating image pixel arrays.
According to my knowledge they should using the same algorithm, hence the speed should not differ much. However my code using interpolate 2D.vi is much slower than Image Resample, especially for larger resample pixels.
On the other hand, the result image I got is strange. I guess the reason is that I am interpolating the U32 pixel value (RGB combied), so should I interpolate RGB values seperately? It will make my code more slow.
Can anyone help me with this? Thanks.
I attach the diagram for my code, in my first code, the hsv to rgb u32.vi is my self-written vi replacing IMAQ color to RGB vi and color value to integer vi.
This self-written has no speed or result problem. The second code is nomal code using NI's vis.
The documentation on the IMAQ Resample VI does not specifically mention what algorithm is used, so it would be difficult to compare it to the Interpolate 2D VI. They also have slightly different interpolation methods that the user can select that could change the results.
Since the IMAQ Resample Help specifically mentions that it accepts RGB image types, I'm assuming it should process color images without any additional code. Would you mind posting your resulting images using the IMAQ Resample and your self-written code?