Depending on how your algorithm works, you can simply concatenate the images in one of the existing image sizes instead of trying to create a ā3Dā image. For example, if you had two frames with 640x480 images, make one 640x960 image as the image to be sent to the GPU.
source
share