There’s a car race going on right now, but it’s not on any sort of race track. There’s a number of companies vying to get their prototype on the road first. [Anurag] has already completed the task, however, except his car and road are functional models.
While his car isn’t quite as involved as the Google self driving car, and it doesn’t have to deal with pedestrians and other active obstacles, it does use a computer and various sensors to make decisions about how to drive. A Raspberry Pi 2 takes the wheel in this build, taking input from a Pi camera and an ultrasonic distance sensor. The Pi communicates to another computer over WiFi, where a neural network operates to make decisions about how to drive the car. It also makes decisions based on a database of pictures of the track, so it has a point of reference to go by.
The video of the car in action is worth a look. It’s not perfect, but it’s quite an accomplishment for this type of project. The possibility that self-driving car models could drive around model sets like model railroad hobbyists create is intriguing. Of course, this isn’t [Anurag]’s first lap around the block. He’s already been featured for building a car that can drive based on hand gestures. We’re looking forward to when he can collide with model busses.
[Daniel Reetz] spent six years working as a Disney engineer during the day and on his book scanner, the archivist at night. Some time last year, [Daniel] decided enough is enough, got married, and retired from the book scanner business. There’s a bit more to it than that, but before leaving he decided to dump, not just the design, but the entire rationale behind the design into a twenty-two thousand word document.
One of his big theses in this document, is his perceived failure of the open hardware movement. The licenses aren’t adequate, as they are based on copyright law that only applies to software. This makes it impossible to enforce in practice, which is why he released the entire design as public domain. He also feels that open hardware shares design, but not rationale. In his mind this is useless when encouraging improvement, and we tend to agree. In the end rationale is the useful thing, or the source code, behind a design that truly matters. So, putting his money time where his mouth is, he wrote down the rationale behind his scanner.
The rationale contains a lot of interesting things. At a first glance the book scanner almost seems a simple design, not the culmination of so much work. Though, once we began to read through his document, we began to understand why he made the choices he did. There’s so much to getting a good scan without destroying the book. For example, one needs a light that doesn’t lose any color information. It doesn’t have to be perfect, as the software can correct the white balance. However, it can’t lean too far away from the natural spectrum, it can’t be too bright, and it can’t be uneven, and it can’t be prohibitively expensive. A lot of thought went into the tent light design.
[Daniel]’s book scanners are immensely popular, and are being used all over the world. He’s certainly made an impact, and the community that formed around his project continues to grow without him. He made some interesting points, and if anything wrote a really good build and design log for the rest of us to learn from.
Pill cameras, devices for ‘capsule endoscopy’, or in much cruder terms, ‘poopable cameras’, are exceedingly cool technology. They’re astonishingly small, communicate through a gastrointestinal tract to the outside world, and have FDA certification. These three facts also mean pill cameras are incredible expensive, but that doesn’t mean a hardware hacker can’t build their own, and that’s exactly what [friarbayliff] is doing for his entry into The Hackaday Prize.
First things first: [friarbayliff] is not building one of these for human consumption. That’s a morass of regulatory requirements and ethical issues. This pill camera is only being built as an experiment, because it would be fun to build one. The pill cams swallowed by patients every day have millions of dollars in R&D behind them before human trials. That said, given a good food-safe enclosure, I’d down one of these as an experiment.
This pill camera will use a simple, off-the-shelf 2 megapixel image sensor that can be bought on eBay for less than five dollars. With a small 32-bit micro, these cameras are easy to drive and capture images from. Power is provided from a single silver oxide button cell battery and a boost converter. In total, [friarbayliff] estimates the total PCB area to be just under one square inch, making this a relatively inexpensive device to build. There will be a radio transceiver in there somewhere, but [friar] hasn’t figure that part out yet.
Pill cameras are some amazing technology, but relatively inaccessible unless you get a used one. Ew. [Mike Harrison] tore one of these pill cams apart a few years ago, and it really is an incredible device. Building one for fun – even if it won’t be used in a human – is a fantastic learning experience and a great entry for the 2016 Hackaday Prize.
A camera slider is an accessory that can really make a shot. But when your business is photography rather than building camera accessories, quick-and-dirty solutions often have to suffice. Thus the genesis of this camera slider controller.
The photographer in question in [Paulo Renato], and while his passion may be photography, he seems to have a flair for motorized dollies and sliders. This controller is a variable-speed, reversible, PIC-based design that drives an eBay gearmotor. The circuit lives on a scrap of perfboard, and it along with batteries and a buck converter are stuffed into the case-modded remains of an old KVM switch. Push buttons salvaged from another bit of e-waste act as limit switches, and a little code provides the magic. We like the hacked nature of the controller, but we wonder about the wisdom of using the former KVM’s USB ports to connect the controller to the drivetrain; it’s all fun and games until you plug a real USB device into it. In sum, though, a nice build with nice results. Check out his other videos for more on the mechanicals.
While the official history of the digital camera begins with a Kodak engineer tinkering around with digital electronics in 1975, the first digital camera was actually built a few months prior. At the Vintage Computer Festival East, [William Sudbrink] rebuilt the first digital camera. It’s wasn’t particularly hard, either: it was a project on the cover of Popular Electronics in February, 1975.
[William]’s exhibit, Cromemco Accessories: Cyclops & Dazzler is a demonstration of the greatest graphics cards you could buy for S-100 systems and a very rare, very weird solid-state TV camera. Introduced in the February, 1975 issue of Popular Electronics, the Cyclops was the first digital camera. This wasn’t a device that used a CCD or a normal image sensor. The image sensor in the Cyclops was a 1 kilobit DRAM from MOS, producing a digital image thirty-two pixels square.
The full description, schematic, circuit layout, and theory of operation are laid out in the Popular Electronics article; all [William] had to do was etch a PCB and source the components. The key part – a one kilobit MOS DRAM in a metal can package, carefully decapsulated – had a date code of 1976, but that is the newest component in the rebuild of this classic circuit.
To turn this DRAM into digital camera, the circuit sweeps across the rows and columns of the DRAM array, turning the charge of each cell into an analog output. This isn’t a black or white camera; there’s gray in there, or green if you connect it to an oscilloscope.
This project in Popular Electronics would be manufactured by Cromemco in late 1975 and was released as their first product in January, 1976. The Cromemco was marketed as a digital camera, designed to interface with the MITS Altair 8800 computer, allowing anyone to save digital images to disk. This was the first digital camera invented, and the first digital camera sold to consumers. It’s an amazing piece of history, and very happy [William] was able to piece this together and bring it out to the Vintage Computer Festival this weekend.
There are a number of ways to control an automobile without using the pedals, and sometimes even without using the steering wheel. Most commonly these alternative control mechanisms are installed in vehicles whose owners are disabled in some way, but [Anurag] has taken this idea of alternative control one step further. He has built a car that can be driven by hand gestures alone.
On a remote controlled car, a Raspberry Pi 2 was installed that handles processing and communication. A wireless network is created on the Pi, and a laptop connects to the Pi over the network. The web camera on the laptop regularly captures frames at 15 fps to check for the driver’s hand gestures. The image is converted to gray scale, thresholded, contours are obtained, and the centroid and farthest points are obtained.
After some calculations are done, a movement decision is taken. The decision is passed to the Pi, which in turn, passed that to the internal chip of the car. All of the code is available on the project’s github page. [Anurag] hopes that this can be scaled up to full sized cars in the future. We’ve seen gesture-based remote controls before that rely on Sonar sensors, so it’s interesting to see one that relies strictly on image processing.
If you are a lover of the aesthetic of vintage photography and Instagram’s filters don’t quite cut it for you, then there are plenty of opportunities even in this post-film age to sample the real thing. Plastic lens cameras from the former Soviet Bloc countries or the Pacific rim are still in production, and you can still buy 35mm and 120 roll film to put in them.
You can even still buy 8mm film for your vintage movie camera, but it’s rather pricey. [Claire Wright] is a young film maker who had an old 8mm camera and really wanted that analog film feel to her work, and she and her father solved this problem by using the 8mm camera’s lens in front of a Raspberry Pi camera sensor. Since an 8mm film frame is 4.5mm x 3.3mm and the Pi camera sensor size is 3.76mm x 2.74mm, it’s quite a good fit.
Their first prototype had a custom case which concealed the Pi camera behind the lens on rails taken from an old CD-ROM drive, and had an HDMI screen on top and a pistol grip to make it portable. An external thumb screw allowed the camera to be positioned in the focal plane.