Oh yeah, this is very important. 3D scanner notes continued. So, you were asking this question. Yes, as noted above, you learned about some practical designs for cheap scanners, but you’ve also noted that those cheap scanners don’t get very good resolution. Now you’re asking the question. How do you get a scanner that is both cheap and high quality?
Okay, here’s one approach that works very well. You attach your laser line generator to the carriage inside of a flatbed scanner. To keep the cost down, it is very simple and cheap to get an old unwanted flatbed scanner that you can modify. If you suppose that your flatbed scanner only scans one line at a time and needs to step the carriage to scan the very next line, and if your flatbed scanner is capable of scanning at a resolution of 1200 DPI, then you know that the stepper motor that controls the carriage is capable of stepping in increments of 1/1200 inch, or 21.167 micrometers. Pretty good resolution, eh? In other words, this is about 1/50 of a millimeter.
BUT… you complain, this requires mechanical moving parts! Aren’t you saying that modern electronics have been increasingly moving away from mechanical moving parts? “Zero moving parts” is the trend in high-tech computer and information technology equipment. Yes, yes, I know, zero moving parts is the cool new spiffy trend, but look. Hard disk drives still have a much higher information storage density and much better storage economy than solid-state drives. So, although the “zero-moving parts” solutions may be trendy, the fact of the matter is, at present, they are not quite practical to complete deprecate systems that have some moving parts.
Okay, but anyways, is it possible to design alternative systems that have no moving parts? Yes, it is. Consider the projector display solution. Rather than moving the laser line, you can just change the image that is displayed to the projector. This makes it very simple to step the projected line in small increments, provided that your projection display has adequate resolution. I recommend using a projector with at least “2K” resolution. Unfortunately, in the realm of projectors, such 2K projectors have only very recently (as of 2016) become economically viable to manufacture in mass numbers, and even then buying one is considerably expensive, measuring in the hundreds of dollars. (And remember, because they are so recent, it’s pretty hard to get an “old” one at a discount.) On the other hand, 2K digital cameras have existed since about 2003. And today (again, as of 2016), 2K digital cameras have gotten considerably smaller and cheaper, down to the price of $25. So, I think it is clear that we have a general trend such that high-resolution sensor equipment is cheaper than high-resolution production equipment.
That being said, one would hope to offload the logic of high resolution capture into the camera system, if possible. Try this idea, for example. Rather than having a precision motion system for your line projector, you just use a cheap $3 laser line generator and move it by hand. Then, you use a high-speed camera to compensate for the fast motion and still get a scan with a high line density regardless. Unfortunately, in the present, the cheapest such high-speed cameras still cost a few thousand dollars.
BUT… now this is really tricky. If you’ve been keeping up on the latest technological developments, you’ll know that delta-image based high-speed cameras are a forthcoming development, and with them will come the capability to build a 10,000 Hz camera at only the cost of a 60 Hz camera. Wow, that’s really impressive. And, considering the fact that we will prefer to keep the camera stationary while moving the laser, the number of image updates will be constrained to only the small area of the image where the laser line is moving, which will dramatically decrease the upper limit on image bandwidth between the camera and the computer system. The only challenge in implementing this system, beyond the fact that the related technology does not yet exist, is writing the software to take in delta image updates and use those to generate the 3D data. However, once all the parts of this system are in place, I think this system will be by far the cheapest, fastest, and highest quality system of all of options mentioned above.
20160803/http://spectrum.ieee.org/at-work/start-ups/startup-profile-sanstreak-lowers-the-cost-of-highspeed-photography
Cottage industry? Interesting.
20160803/https://en.wikipedia.org/wiki/Putting-out_system
20160803/https://en.wikipedia.org/wiki/Ironmonger
3D scanner notes continued continued. What about the kind of camera you want to use? Look, I think I want to make this point clear. First of all, get a camera that has some kind of image focus ability. Smartphone cameras that have no image focus abilities are notorious for giving you out-of-focus images for near-range photographs. It’s absolutely necessary to have a focus mechanism if you are photographing small objects. This might mean that you want to go for a full macro lens from a DSLR camera.
Oh, but this is interesting. Not all smartphone cameras are poor at focusing. Yeah, and I’ve picked up a few other entertaining links in the process, though they are not as informative to the primary objective here.
20160911/http://petapixel.com/2016/09/07/iphone-7-7-plus-announced-dual-cameras-zoom-bokeh/ 20160911/http://petapixel.com/2016/09/08/iphone-7-simulate-shallow-depth-field/
Wow, now I must say this is interesting. Yes, definitely as the author of the article has said, there are not many electronics in this newer camera. The space is not packed and busy like the rhetoric of older cameras. It doesn’t quite have as much of that echo of a sophisticated industrial product.
20160911/http://petapixel.com/2016/09/09/teardown-take-peek-inside-canon-5d-mark-iv/
20160911/http://petapixel.com/2016/09/09/photographer-makes-website-seeking-sugarmama-fund-adventures/
20160911/http://petapixel.com/2016/09/07/patent-shows-canons-digital-lens-displays-may-look-like/
20160911/http://petapixel.com/2016/09/08/shoot-portrait-levitating/
Very important. This is the site of the department at the University of Minnesota that is likely to be the one to go to to get your 3D scanner ideas moved forward.
20160914/http://www.dtc.umn.edu/