what is the purpouse of uint 8 in image processing?

am very new to this area., can anyone tell me please what is uint8 ,16,double in image processing why we use this?

Answer 1

[Image Analyst]

That's the data type. All variable have a data type of course. In image processing you can use anything, and the ones you listed are particularly common, especially uint8 for images stored in disk files. You don't really "use" these - they're simply a data type. I'm not sure what else to say.

Answer 2

[Walter Roberson]

Usually with color images, 8 bits of information per channel (Red, Green, Blue) are stored. That is a total of 24 bits of information per channel. The human eye has difficulty distinguishing so many tones, especially in typical home or office lighting environments, so 8 bits per channel is often enough for typical photographs.

However, in some cases, the various channels are not intended to represent typical photographic color information, and might instead represent value readings from precision instruments, such as multi-band infrared scans looking for various underground mineral formations (or checking for fires, or measuring ocean temperatures, etc.) In such case it is useful to be able to have more than 8 bits per channel.

Higher resolution is also often used for astronomical photographs, where the hidden details might be beyond what the eye can normally perceive (in in frequencies not visible to humans.)

Why not less than 8 bits? Because especially in some colors, it isn't hard for a human to pick out more than 64 shades (6 bits), and although humans aren't great beyond 128 shades, the trouble of packing down 7 bits per channel in to standard 8-bit bytes just isn't worthwhile.