OAK-D Depth Sensing AI Camera Gets Smaller And Lighter

The OAK-D is an open-source, full-color depth sensing camera with embedded AI capabilities, and there is now a crowdfunding campaign for a newer, lighter version called the OAK-D Lite. The new model does everything the previous one could do, combining machine vision with stereo depth sensing and an ability to run highly complex image processing tasks all on-board, freeing the host from any of the overhead involved.

Animated face with small blue dots as 3D feature markers.
An example of real-time feature tracking, now in 3D thanks to integrated depth sensing.

The OAK-D Lite camera is actually several elements together in one package: a full-color 4K camera, two greyscale cameras for stereo depth sensing, and onboard AI machine vision processing with Intel’s Movidius Myriad X processor. Tying it all together is an open-source software platform called DepthAI that wraps the camera’s functions and capabilities together into a unified whole.

The goal is to give embedded systems access to human-like visual perception in real-time, which at its core means detecting things, and identifying where they are in physical space. It does this with a combination of traditional machine vision functions (like edge detection and perspective correction), depth sensing, and the ability to plug in pre-trained convolutional neural network (CNN) models for complex tasks like object classification, pose estimation, or hand tracking in real-time.

So how is it used? Practically speaking, the OAK-D Lite is a USB device intended to be plugged into a host (running any OS), and the team has put a lot of work into making it as easy as possible. With the help of a downloadable application, the hardware can be up and running with examples in about half a minute. Integrating the device into other projects or products can be done in Python with the help of the DepthAI SDK, which provides functionality with minimal coding and configuration (and for more advanced users, there is also a full API for low-level access). Since the vision processing is all done on-board, even a Raspberry Pi Zero can be used effectively as a host.

There’s one more thing that improves the ease-of-use situation, and that’s the fact that support for the OAK-D Lite (as well as the previous OAK-D) has been added to a software suite called the Cortic Edge Platform (CEP). CEP is a block-based visual coding system that runs on a Raspberry Pi, and is aimed at anyone who wants to rapidly prototype with AI tools in a primarily visual interface, providing yet another way to glue a project together.

Earlier this year we saw the OAK-D used in a system to visually identify weeds and estimate biomass in agriculture, and it’s exciting to see a new model being released. If you’re interested, the OAK-D Lite is available at a considerable discount during the Kickstarter campaign.

3D Printed Adapter Puts Slides In Their Best Light

If you’ve got old family photos on slides there’s an excellent chance you’ve considered digitizing them at one point or another, but perhaps didn’t know the best way of going about it. In that case, this 3D printed adapter designed by [Rostislav Persion] that lets you photograph slides with a standard DSLR may be exactly what you were waiting for.

The idea is simple enough, you place the slide inside the adapter, get your focus right, and snap a picture. But of course, you’ve also got to provide some illumination. In this case, the camera is mounted on a tripod and pointed at an appropriate light source. Once you’ve experimented a bit and got the image backlit the way you want it, you can lock everything in place and easily power through a stack of vintage family memories in no time.

For such a straightforward concept, we really appreciate the little details in the execution. For example, rather than just sliding a 3D printed cylinder over the DSLR’s lens, [Rostislav] came up with a foam-padded “shim” that’s strong enough to hold the adapter on without marring anything. The two-part slide spacer that features a bit of springiness to hold everything tight is also a very nice touch.

An approach like this should work nicely for the amount of slides most families are likely to have, but if you’re in a position where you need to digitize thousands of images, some automation would certainly help things along.

World’s Cheapest And Possibly Worst IR Camera

Don’t blame us for the title. [CCrome] admits it may well be the cheapest and worst IR camera available. The concept is surprisingly simple. Mount a cheap Harbor Freight non-contact thermometer on a 3D printer carriage and use it to scan the target. The design files are available on GitHub.

There is, of course, an Arduino to grab the data and send it to the PC. Some Python code takes care of converting it into an image.

Perhaps you don’t need a camera, but having a way to communicate with an $11 IR temperature sensor might come in handy someday. You do have to mash the measurement button down, so [CCrome] used the 3D printer to make a clamp for the button that also holds the POGO pins to the PCB. We would have been tempted to solder across the switch and also solder the wires to the pad. But, then again, you need a 3D printer for the project anyway.

Don’t expect the results you would get from a real thermal sensor. If you want that, you may have to build it yourself or open your wallet wide. If you need some inspiration for a use case, look at the thermal camera contest from a few years back.

Mini Camera Crane For Your Workbench

If you’ve ever tried to document a project on your workbench with photos or videos, you know the challenge of constantly moving tripods to get the right shot. [Mechanistic] is familiar with this frustration, so he built a small desktop camera crane.

Heavily inspired by [Ivan Miranda]’s large camera crane, this build scales it down and mainly uses 3D printed parts. The arm of the crane can pivot along two axes around the base, uses a parallel bar mechanism to keep the camera orientation constant through its vertical range of motion. The camera mount itself allows an additional 3 degrees of freedom to capture any angle and can mount a DSLR or smartphone. To offset the weight of the camera, an adjustable counterweight is added to the rear of the arm. Every axis of rotation can be locked using thumbscrews.

We can certainly see a crane like this being useful on our workbench for more than just camera work. You could create attachments for holding lights, displays, multimeters, or some helping hands. For some tips on creating an engaging project video check out [Lewin Day]’s excellent video on the subject.

Continue reading “Mini Camera Crane For Your Workbench”

Printed Adapter Puts Vintage Lens Back To Work

While browsing through an antiques shop, [Nick Morganti] came across a Kodak slide projector with an absolutely massive lens hanging off the front. Nearly a foot long and with a front diameter of approximately four inches, the German-made ISCO optic was a steal for just $10. The only tricky part was figuring out how to use it on a modern DSLR camera.

After liberating the lens from the projector, [Nick] noted the rear seemed to be nearly the same diameter as the threaded M42 mount that was popular with older film cameras. As luck would have it, he already had an adapter that let him use an old Soviet M42 lens on his camera. The thread pitch didn’t match at all, but by holding the lens up to the adapter he was able to experiment a bit with the focus and take some test shots.

Encouraged by these early tests, [Nick] went about designing a 3D printed adapter. His first attempt was little more than a pair of concentric cylinders, and was focused like an old handheld spyglass. This worked, but it was quite finicky to use with the already ungainly lens. His second attempt added internal threads to the mix, which allowed him to more easily control focus. After he was satisfied with the design, he glued a small ring over the adapter so the lens could no longer be unscrewed all the way and accidentally fall out.

To us, this project is a perfect application of desktop 3D printing.[Nick] was able to conceptualize a one-of-a-kind design, test it, iterate on it, and arrive on a finished product, all without having to leave the comfort of his own home. To say nothing of the complex design of the adapter, which would be exceedingly difficult to produce via traditional means. Perhaps some people’s idea of a good time is trying to whittle a lens bayonet out of wood, but it certainly isn’t ours.

So it’s probably little surprise we’ve seen a number of similar projects over the years. From monstrous anamorphic adapters to upgraded optics for the Game Boy Camera, it seems there’s a healthy overlap between the 3D printing and photography communities.

Mastering Stop Motion Through Machine Learning

Stop motion animation is notoriously difficult to pull off well, in large part because it’s a mind-numbingly slow process. Each frame in the final video is a separate photograph, and for each one of those, the characters and props need to be moved the appropriate amount so that the final result looks smooth. You don’t even want to know how long Ben Wyatt spent working on Requiem for a Tuesday, though to be fair, it might still get done before the next Avatar.

But [Nick Bild] thinks his latest project might be able to improve on the classic technique with a dash of artificial intelligence provided by a Jetson Xavier NX. Basically, the Jetson watches the live feed from the camera, and using a hand pose detection model, waits until there’s no human hand in the frame. Once the coast is clear, it takes a shot and then goes back to waiting for the next hands-free opportunity. With the photographs being taken automatically, you’re free to focus on getting your characters moving around in a convincing way.

If it’s still not clicking for you, check out the video below. [Nick] first shows the raw unedited video, which primarily consists of him moving three LEGO figures around, and then the final product produced by his system. All the images of him fiddling with the scene have been automatically trimmed, leaving behind a short animated clip of the characters moving on their own.

Now don’t be fooled, it’s still going to take awhile. By our count, it took two solid minutes of moving around Minifigs to produce just a few seconds of animation. So while we can say its a quicker pace than with traditional stop motion production, it certainly isn’t fast.

Machine learning isn’t the only modern technology that can simplify stop motion production. We’ve seen a few examples of using 3D printed objects instead of manually-adjusted figures. It still takes a long time to print, and of course it eats up a ton of filament, but the mechanical precision of the printed scenes makes for a very clean final result.

Continue reading “Mastering Stop Motion Through Machine Learning”

Automatic Microfiche Scanner Digitizes Docs

While the concept might seem quaint to us today, microfiche was once a very compelling way to store and distribute documents. By optically shrinking them down to just a few percent of their original size, hundreds of pages could be stored on a piece of high-resolution film. A box of said films could store the equivalent of several gigabytes of text and images, and reading them back only required a relatively simple projection machine.

As [Joerg Hoppe] explains in the write-up for his automatic microfiche scanner, companies such as Digital Equipment Corporation (DEC) made extensive use of this technology to distribute manuals, schematics, and even source code to their service departments in the 70s and 80s. Luckily, that means hard copies of all this valuable information still exist in excellent condition decades after DEC published it. The downside, of course, is that microfiche viewers aren’t exactly something you can pick up at the local Big Box electronics store these days. To make this information accessible to current and future generations, it needs to be digitized.

The camera panning over a full DEC microfiche sheet.

[Joerg] notes there are commercial services that would do this for you, but the prices are just too high to be practical for the hobbyist. The same for turn-key microfiche scanners. Which is why he’s developed this hardware and software system specifically to digitize DEC documents. The user enters in the information written on the top of the microfiche into the software, and then places it onto the machine itself which is based on a cheap 3D printer.

The device moves a Canon DSLR camera and appropriate magnifying optics in two dimensions over the film, using the Z axis to fine-tune the focus, and then commands the camera to take an image of each page. These are then passed through various filters to clean up the image, and compiled into PDFs that can be easily viewed on modern hardware. The digital documents can be further run though optical character recognition (OCR) so the text can be easily searched and manipulated. In the video after the break you can see that the whole process is rather involved, but once the settled into the workflow, [Joerg] says his scanner can digitize 100 pages in around 10 minutes.

A machine like this is invaluable if you’ve got a trove of microfiche documents to get through, but if you’ve just got a sheet or two you’d like to take a peek at, [CuriousMarc] put together a simple rig using a digital microscope and a salvaged light box that should work in a pinch.

Continue reading “Automatic Microfiche Scanner Digitizes Docs”