Okay, so the output is a 24b RGB image. We can assume this means an 8,8,8,0,0 arrangement; that's pretty common. So the goal seems specific.
The input is "any 8 bit, 16 bit, 32 bit, 64 bit image". It's not enough to say that's unclear. It's universal. Are those also RGB formats? Maybe. Are they I/IA/RGBA/indexed? Maybe. How are the contents of an RGB pixel packed into 16 bits? Any way they might be. Are these images represented as a prismatic arrays in the workspace, or are they still somehow encoded in a file on disk? Maybe. Are they being read via imread()? What about fread()? -- or are they being accessed via .NET? Yes, maybe. Is any of this design ambiguity resolved prior to the developed solution being applied? No, apparently not.
What is the route from anywhere to Albequerque NM? It's the envelope which contains both Albequerque and the entire universe. Allot extra travel time in case it turns out that you didn't actually even want to go to Albequerque in the first place.
