Okay, so now we have some rather interesting ideas to integrate with how we design our 3D scanner.
First of all, just to get the quick one out of the way, our naming for our “offset map.” To make this more of a coherent metaphor with the photography profession, we can call this “circles of confusion.” After all, that is what we call the artifacts seen in photographs that are out of focus. Anyways, yes, the map that shows the size of the same-time-colored dot in relation to the position of the pixel.
Second, on thresholding. In addition to recording the ambient light photograph to compare against for thresholding laser light, we can also use the color of the pixel in the ambient light map to determine the reflectivity of the material at that point to the laser light. For example, bright red or white will be very reflective to our red laser light, while blue surfaces are going to need more sensitive thresholding. So in the case of the bright red or white surfaces, we can adjust our algorithms to use the least sensitive thresholding, or the highest thresholding in other words. The purpose of this is to reduce our margin of error in detecting whether a pixel is very closest to the center of our laser, thus opting for the minimum “circle of confusion” and getting the highest resolution.
Now, for the third new idea. On detecting the position of the laser. There is a very ingenious technique that we can use for this. It is almost too simple: shadowing. So, first of all, we have our known scanning structure that is computer printed and can be used to triangulate the position of the camera based off of the dot positions in the photograph. Now we shine our laser line generator on this structure. We have this particular structure setup so that there are two white paper poles in front of the flat white cardboard background. This will cause a laser to appear on the background cardboard, with visible shadows caused by the poles, and laser light visibly illuminating some positions on the poles. This can be used to draw two lines through the shadows to the illuminated positions on the poles that point us directly to the laser’s position. Now that we know the exact position of the laser, we can use our laser profile map to map the exact directions of the laser rays, in the case that our laser does not project a perfectly straight line, and for the sake of laser intensity thresholding.
Also, we can rotate the laser around its axis a little bit and do a few more such exposures to determine our axis of rotation and even our laser module’s offset from the axis of rotation.
Finally, we can use our recognized background during the course of the scan to calibrate the laser’s angular position in relation to time. And we don’t need to know all of this other information during the scan, which means we can remove the poles from the set during the scan.