Every day your eyeballs are assaulted by advertisements on your box of cereal, billboards, t-shirts, magazines, milk cartons, plastered on the side of buses, buildings, bananas, and written in the sky. [Reed], [Jonathan], [Tom], and [Alex] came up with a solution to this: a Brand Killer that censors all the advertisements and brands you see every minute of every day. It’s a real-world adblock that you can build right now.
The team’s system uses a custom head mounted display made from cardboard, goggles, a webcam, and a seven-inch display. The software for the system uses Python and OpenCV to monitor the images from the webcam, compares them against a list of brands and logos, and filters them out with an unobtrusive blur.
Right now the system just has a few brands and logos that include Dr. Pepper, Hershey’s, McDonalds, Facebook, Starbucks, and clear evidence this was built at UPenn, Wawa and Tastykake. In the video below, the detection and tracking of these various brands is very good. The system is also stereoscopic, meaning this is wearable all day, every day, without a loss of depth perception.
Continue reading “Real World AdBlock”
[Artlav] wanted to build a digital camera, but CCDs are expensive and don’t respond well to all wavelengths of light. No problem, then, because with a photodiode, a few stepper motors, the obligatory Arduino, and a cardboard box, it’s pretty easy to make one from scratch.
The camera’s design is based on a camera obscura – a big box with a pinhole in one side. This is all a camera really needs as far as optics go, but then there’s the issue of digitizing the faint image projected onto the rear of the camera. That’s fine, just build a cartesian robot inside the box and throw a photodiode in there.
There are a few considerations when choosing a photodiode for a digital camera. Larger photodieodes have higher noise but lower resolution. [Artlav] has been experimenting with a few diodes, but his options are limited by export control restrictions.
Even with the right photodiode, amplifying the tiny amount of current – picoamps in some cases – is hard. The circuit is extremely sensitive to EMI, and it’s inside a box with stepper motors pulled from the scrap bin. It’s amazing this thing works at all.
Still, [Artlav] was able to get some very high resolution images across a huge range of wavelengths. He’s even getting a few images in mid-wave infrared, turning this homebrew digital camera into the slowest thermal imaging camera we’ve ever seen.
There are certainly battery hungry devices out there on the market and, unless you do some serious research before the purchase of said device, you really don’t know how it will perform. Needless to say, some of us get stuck with power hog device, and it seriously sucks because changing out batteries often is expensive and just plain annoying.
If you couldn’t tell, I am speaking from experience, my old Sony DSC-H5 camera works great with the exception of needing new batteries every hour. And if you get cheap batteries, the camera won’t even turn on! There’s a USB connector on the camera but it is only for transferring data and there is also no DC input jack. The entire situation is a totally bummer.
I’m happy (or disappointed) that I am not alone in the world. [Phil] wrote into the HaD tip line to tell us about his solution to this very problem. He has a Canon SD1000 camera and although the battery works fine he needs it to work at an altitude of 15km in order to take some sunrise photos. Cold weather testing (in the fridge freezer) showed that the battery isn’t going to cut the mustard for the hour-long flight. The rest of the balloon-lifted unit already has a battery pack and the plan would be to tap into that to power the camera. Unfortunately his camera, like mine, doesn’t have a DC input jack and can not be powered off the USB port.
[Phil] decided to make a 3D printed battery emulator. It sits in place of the stock battery and holds bare wire where the batteries terminals normally are. The other end of the wires are run out of the camera to a voltage regulator that converts the battery pack’s 6 volts down to the 3.9 that the camera needs.
Continue reading “3D Printed Camera Battery Emulator”
[Glitchmaker] loves photography and wrote in to tell us about his newest project. He has a Canon 1000D camera but, unfortunately, it does not have time lapse capability. So, instead of shelling out a chunk of change for a new camera [Glitchmaker] decided to make an external shutter control device that can continue to instruct the camera to take photos at predetermined intervals. He calls his project: SHTTTRRR. You didn’t think that meant something else, did you?
You can see the unassuming box above, there is just enough stuff packed in there to get the job done, nothing extra or fancy. Luckily, the Cannon camera has a remote shutter input jack that only requires connecting one pin to another in order to take a photo. Inside the box is an ATTINY45 microcontroller. It reads the button pushes from the single panel-mounted button and calculates the time between two button presses. That time between button presses determines the frequency of the photos taken. At the appropriate times, the ATTINY45 signals a transistor to connect the two appropriate pins on the camera’s remote shutter input jack. The device continues to tell the camera to take photos until it is shut off. The result is a series of time-lapse photos that was previously not possible on that camera!
This is a simple project that solves a problem and gets the job done. What’s better than that? [Glitchmaker] is proud of the SHTTTRRR he made and also learned a bunch about programming the ATTINY45 along the way. Check a video of it working after the break.
Continue reading “Full SHTTTRRR Control Lets You Take Your Time…”
[Tesla500] has a passion for high-speed photography. Unfortunately, costs for high-speed video cameras like the Phantom Flex run into the tens or even hundreds of thousands of dollars. When tools are too expensive, you do the only thing you can – you build your own! [Tesla500’s] HSC768 is named for the data transfer rate of its image sensor. 768 megapixels per second translates to about 960MB/s due to the 10 bit pixel format used by the On Semiconductor Lupa1300-2 image sensor.
This is actually [Tesla500’s] second high-speed camera, the first was HSC80, based upon the much slower Lupa300 sensor. HSC80 did work, but it was tied to an FPGA devboard and controlled by a PC. [Tesla500’s] experience really shows in this second effort, as HSC768 is a complete portable system running Linux with a QT based GUI and a touchscreen. A 3D printed case gives the camera that familiar DSLR/MILC shape we’ve all come to know and love.
The processor is a Texas Instruments TMS320DM8148 DaVinci, running TI’s customized build of Linux. The DaVinci controls most of the mundane things like the GUI, trigger I/O, SD card and SATA interfaces. The real magic is the high-speed image acquisition, which is all handled by the FPGA. High-speed image acquisition demands high-speed memory, and a lot of it! Thankfully, desktop computers have given us large, high-speed DDR3 ram modules. However, when it came time to design the camera, [Tesla500] found that neither Xilinx nor Altera had a FPGA under $1000 USD with DDR3 module support. Sure, they will support individual DDR3 chips, but costs are much higher when dealing with chips. Lattice did have a low-cost FPGA with the features [Tesla500] needed, so a Lattice ECP3 series chip went into the camera.
The final result looks well worth all the effort [Tesla500] has put into this project. The HSC768 is capable of taking SXGA (1280×1024) videos at 500 frames per second, or 800×600 gray·scale images at the 1200 frames per second. Lower resolutions allow for even higher frame rates. [Tesla500] has even used the camera to analyze a strange air oscillation he was having in his pneumatic hand dryer. Click past the break for an overview video of the camera, and the hand dryer video. Both contain some stunning high-speed sequences!
Continue reading “[Tesla500] Builds a High-Speed Video Camera”
Professor [Bruce Land] teaches a microcontroller class at Cornell University, and it seems like this year’s theme was selfie-taking-robots.
First up is a clever mix of technology by [Han, Bihan and Chuan]. What happens when you take an iPhone, three microphones and a microcontroller? The ultimate device in selfie-taking-technology, that’s what — Clap-on! The iPhone is mounted on a few servo motors which allows the bot to direct the camera towards, you guessed it, a clapping noise. On the second clap, the phone takes your picture. Cute.
Next up is a bit more sophisticated — a facial recognition selfie-bot. This little robot can be programmed to track faces and take pictures of you and your friends when your arm is just not long enough. Not only that, you can set all kinds of parameters so you get the perfect picture. It uses OpenCV to crunch the raw data and outputs commands to an ATmega1284 which controls the servo motors that direct the camera. This project was by [Michael and Jennifer] — two fourth year students at Cornell.
Continue reading “Selfie-Bots Will Take Your Best Shots For You”
All [val3tra] wanted was an RF-accessible camera. A camera that would take pictures, save them to an SD card, and occasionally send them over an RF link to a computer. This project has grown out of control, and now it has become an open-source camera that’s able to take year-long time-lapse movies.
The build started as a low power camera using an eBay JPEG camera modified for 3.3V. That’s only 640×480, but each frame averages only 48kb – small enough to store a few thousand pictures on a FAT16 formatted SD card. A $4 RF module, an ATMega, and an RTC make up the rest of the build that has a power draw of about 100 Joules per hour. A D-cell has about 60,000 Joules, and a pessimistic estimate of a battery of four in series, two in parallel gives a run time of 200 days.
This build was then improved, bringing the total battery consumption down to about 3.5-4 Joules per frame, or at one frame every 10 minutes, about 24 Joules an hour. That’s impressive, and getting this camera to run longer than a dozen or so months raises some interesting challenges. The self-discharge of the battery must be taken into account, and environmental concerns – especially when leaving this camera to run in a Moscow winter, seen in the video below – are significant.
If you don’t want to go equipment-lite you could seal your DSLR, Pi, and some serious batteries in a weatherproof enclosure.
Continue reading “A Year Long Time Lapse Camera”