Here’s yet another example of well targeted advertising. This camera built around a Raspberry Pi is a giveaway from Sprite. The “lucky” winner of the camera will have the pleasure of seeing the Sprite logo as a watermark on all of the images they snap with it. But in the right hands it’s a simple hack to remove that “feature” (they published the Python script that adds the watermark) or to just scrap the parts for another project. Either way, Sprite got us to say their name three times in this paragraph so the campaign worked.
The most obvious part of this build is the custom cast resin case that they came up with which is a gaudy cartoon-like monstrosity. It protects the case-less Raspberry Pi board, and mounts the Pi Camera board so that the lens is positioned correctly. The lipstick-sized module mounted in the lower back half of the case is a 2400 mAh portable power supply with a USB charging port sticking out the side. This makes us wonder, do you have to wait for the RPi to power up before snapping a picture? If the size and color didn’t get you noticed by everyone the shutter sound will. it shouts the name of the soda company whenever you press the shutter release button.
If you’re more of a high-end photography enthusiast this DSLR wedded with an RPi will be of more interest.
A while back our good buddy [Ch00f] built a QR code clock, unreadable to both humans and computers. A human couldn’t read the clock because of the digital nature of a QR code, and because the clock used persistence of vision in driving the LEDs, a digital camera can’t capture all the pixels in the QR code at the same time. It’s a highly useless but impressive art piece. Now, [Ch00f] is turning that build on its head. He created a rudimentary display that is invisible to the human eye, but easily detected with a digital camera.
This build exploits a basic property of CMOS digital cameras – the rolling shutter. Because it takes time to get pixels off a modern digital image sensor, each picture is actual a composite of many different strips, each taken slightly out of sequence. You can see this for yourself by taking a picture of something rotating very fast with your camera phone; a picture of an airplane propeller will make the blades appear curved, or look like [Dr. Seuss] has an aeronautical engineering degree.
To create his display, [Ch00f] found a few inexpensive fiber optic lights. By aligning a few of these into columns and lighting them up in a precise sequence, he can exploit the rolling shutter and make an image appear. To the human eye, it looks like a solid wall of illuminated fiber optics.
As for how practical this build is, [Ch00f] says not much. For cell phone cameras, you’d need to have a very, very short exposure time for this to work. The only way to do that is to make this display unbelievably bright, or just put it out in the sun. We can’t see that being practical for any potential use case, but we’d be more than happy to see a large-scale attempt at displaying images with this technique.
Here’s [Tom Parker] showing off a brushless motor gimbal stabilizer for his GoPro camera. We saw a similar project a couple of weeks back that featured a 3d printed quadcopter mount. This offering is meant to be held in your hands. It keeps the subject in frame even if the cameraman’s hands pitch and roll (we figured aeronautical terms were best here). This image shows him demonstrating a level camera as he quickly rolls the frame from one side to the other. It doesn’t compensate for yaw, which is something he may change in the next iteration. We already like the results he’s getting with it.
About 3:15 into the video demo below we get a very quick description of the build itself. He started it as a project at University. Fabrication included work on a 3D printer, laser cutter, and vacuum forming machine. The grips are bicycle handlebar components. To overcome the stabilization system the operator has access to a joystick. Without this you’d never be able to aim the camera up or down because of auto-leveling.
Continue reading “Showing off a high-performance brushless motor camera gimbal”
Glass work is always a feast for the eyes, especially when it is hot glass. Watch as a Nikkor lens is made from beginning to end. It is wonderful to see the care taken to search by eye for defects, refraction issues, clarity etc. It may just be for the video, but it seems that the workers truly do take pride in their product.
What I found somewhat surprising was the amount of work that went into refining the glass BEFORE it was even put into a lens mold. I would have assumed that much of the work would have come after.
This camera rig uses a Raspberry Pi to send a camera down fifty meters (mirror on RPi blog) in order to spy on sharks. We got really excited at first thinking that it might be using the camera module from the Raspberry Pi Foundation but that isn’t the case. Do keep reading though, there’s a lot of cool stuff involved in this one.
The project used a collection of camera units spread over a large area to monitor shark activity. Each is mounted on an anchored buoy, using solar panels and a lead acid gel battery for power. The RPi itself remains topside in a waterproof box. It connects to the camera using a 50-foot Ethernet patch cable.
We figure the challenge of building the hardware parallels that of designing an underwater ROV. The camera needs an enclosure that can stand up to the pressure at that depth while allowing the cable to pass through it. There is also an interesting note in the project log about getting the camera exposure settings to behave.
[Chonggang Li] wrote in to share a link to the final project he and [Ran Hu] built for their embedded systems class. It’s called Piano Hero and uses an FPGA to implement a camera-based touch screen system.
All of the hardware used in the project is shown above. The monitor acts as the keyboard, using an image produced by the FPGA board to mark the locations of each virtual key. It uses a regular VGA monitor so they needed to find some way to monitor touch inputs. The solution uses a camera mounted above the screen at an obtuse angle. That is to say, the screen is tilted back just a bit which allows the images on it to be seen by the camera. The FPGA board processes the incoming image, registering a key press when your finger passes between the monitor and the camera. This technique limits the input to just a single row of keys.
This should be much simpler than using a CCD scanner sensor, but that one can track two-dimensions of touch input.
Continue reading “Camera-based touchscreen input via an FPGA”
How lucky is [Transistor Man] that he found the materials for the tracks of this curved camera dolly just lying around the shop? The three rails making up the system are quarter-inch diameter and he was able to bend them by hand with the help of a 55 gallon drum. But to hold them in place so that the camera dolly would run smoothly he had to find a way to precisely space the tracks.
The robot arm you see in the picture above is a 3D printer which ended up being the easiest solution to the problem. With a bit of trial and error he found a design that holds the tracks in place without interfering with the camera sled’s progress. From there he devised a mounting system which uses three camera tripods to hold the track. You can see a test video shot from the dolly track embedded after the jump. It’s the opposite of the bullet time rigs [Caleb’s] been working on lately.
We figure the spacers would work for any track shape, but if you’re going for a complicated route you need some type of pipe bender to help out.
Continue reading “Bending and printing a curved camera dolly track”