Help with building RGB LASER Scanner?

there are several ways to do this, each involving different amounts of electro-optics in digital and analog forms.

you'll have to decide what you're scanning. is it 2D? 3D? still? or live-action?

in the easiest of cases, you can probably find an old photocopy machine and see how that works.

in the hardest case, your scanning data might become real-time input to a vehicle that must navigate through its environment.

so you will have a scanning beam, and a detector.

the scanning beam is a laser; you're interested in tri-colour, so this is either three lasers of different wavelengths, or it is a white-light laser that you will later filter. true white-light lasers are pulsed. if the pulse rate is really high (eg: in the millions of pulses/second), then each pulse could represent a pixel, and still have a mean free path of several meters.

the detector is a CCD photoreceptor, (or a whole lot of them), filtered to pass the desired wavelength(s). you can choose to capture:

a) a stream of individual bounces without regard to the focus;

b) a whole scan line into a buffer at once; or

c) a whole 2D frame into a buffer at once.

each method implies different mechano-optics, with a) being the simplest, mechanically. additionally, it may be possible to convert some digital cameras to use for this purpose.

to scan, you'll draw a point of light across a remote object, detecting the intensity of what bounces back. if this is not a range-finding device, then the outbound signal may be constant-intensity; but to detect distance, velocity, or absorption spectra, you will have to do some optical and/or high-speed tricks (sub-nanosecond processing is hard) involving correlating the outbound signal to the returned signal, to build a profile.

photon time-of-flight would indicate distance, (1 ns is about 30cm), and doppler shift would indicate speed.

to detect time of flight implies very high sensitivity and probably high frame rate, too.

to detect doppler shift and spectral lines would require that the returned beam travel through a prism, to give angular separation to the photon stream by wavelength. in this case, you'll always be scanning into a buffer, whose width covers the returned wavelengths.

each scan line would be drawn using a mirror that can tilt and swivel on two orthogonal axes. (or a pair of mirrors, arranged to swivel on orthogonal axes). controlling the mirror(s) will be subtle and delicate, because vibration may cause some fuzz in where the point gets drawn.

(think of a mechanism very similar to a DVD reader, for instance.)

for noise control, you'll want to distinguish between ambient light and returned signal.