The system rotates the coin, sampling its surface using the 30 x 30 pixel resolution optical sensor. Then it compares it to a stored model, giving you the nay or the yay with more accuracy than anyone, or about the same accuracy as an expert. [El Mundo — In Spanish]

While the concept is not new, this prototype is fully working and operational. Created by Jing Li and a team of researches at NASA’s Ames Research centre, Moffett Field, California, the sensor is a multiple-channel silicon-based sensing chip integrated in micro-board with 64 nanosensors.

The low-cost, low-power system can detect minimal concentrations of ammonia, chlorine gas and methane, showing the values in an iPhone application. It can automatically communicate the results with other mobile phones or the Enterprise’s computer using Wi-Fi or 3G, and order massive teleportation evacuations if needed. OK, not true. No teleportation yet, but we are getting there. [NASA]

Let me explain. The same biological structures used in the mantis shrimp—a strange creature with 500m years of lineage—to see circular polarised light are used in modern satellite communications and the laser/lens setup in DVD and CD players. The fundamental hardware in our tech are called quarter-wave plates. They both do a 90-degree shift to wave patterns, and quarter waves are ideal for our tech because they are part of high compression, lossless data signals. But mantis shrimp eyes can work this magic on a broad array of light spectrum with their tightly packed tube shaped cells, while our clumsy human tech can only do it to a few frequencies. The potential for new higher bandwidth devices is there.

[Wired]

The OV5653, based on the company’s latest 1.75-micron OmniBSI technology, delivers low-light sensitivity of >1400mV/(lux-sec) – a 40% improvement over frontside illumination technology – and a two times improvement in signal-to-noise ratio (< 70lux), according to OmniVision. The OV5653 enables 720p recording at 60 fps and full HD 1080p at 30 fps, making it ideal for both 5-megapixel DSCs and DVs, according to the company.

So not only they have achieved 720p 60fps and 1080p at 30fps, but their OV5653 and OV5650 parts also have better sensibility and two times less signal-to-noise ration. Not bad at all. [Digitimes via Into Mobile]

At All Things D, my friends at Fullpower did a demo of a accelerometer-equipped headset that can pick up a call by tapping it in a different way than you tap a regular headset.

That’s not so new in concept, but the trick is, they use math to filter out the background noise—in this case, motion from walking, jumping, etc—so it doesn’t hang up on you when you move around while doing it.

The headset also knows when you place it on a table and powers down. All by using math and a regular accelerometer.

You’ve seen games and GPS apps from Fullpower but those are just apps demoing the company’s tech. Fullpower’s motion detection engine tech is described as doing for motion what voice recognition does for voice. It interprets the raw data and figures out what a person is doing, eliminating confusing data, which I think is interesting because up to now, most developers have just had to deal with raw accelerometer XYZ information. Hard to parse in itself, but up to now, really hard to take that info and decipher what exactly the person holding the device is doing.

Next up is an AMAZING demo of a camera app that filters out motion using the accelerometer. Typically, software that have done this has done it by using gyroscopes, or mechanical parts, or by digitally scanning the image as you move it. The accelerometer here helped the camera, mounted on a wildly shaking platform. The images are taken on a crappy smartphone sensor (a slow sensor), came out very sharp when stabilisation is applied. I’m unsure if it’s timing it properly to snap when the motion is at its slowest, but that would make sense, since there’s no way to increase shutter speed. The tech can scale to all sorts of high end cameras, using just cheap accelerometer parts, not the typically high end stuff you see in DSLRs now. I look forward to getting this stuff in smartphones.


The demos were just concepts, but I’m sure we’ll see more of this tech in products, soon. [Fullpower]

[Disclosure: these guys are my friends.]

The metal detector is outfitted with a relatively inexpensive “combustible gas sensor” that can detect pockets of oil, and the creator notes that the entire mod can be completed for around $US100. This particular oil detector includes some green LEDs which we don’t believe serve any other purpose than looking cool (as we deduced from the intro, “What good is your newly built metal detector without a little bling?”).

Neighborhoods in which this mod is guaranteed to work include the Mission District in San Francisco, Wicker Park in Chicago, and any Brooklyn neighbourhood primarily inhabited by white people.

Disclosure: Your humble editor also has a beard. And square, plastic-rimmed glasses. This post is thus what is known as “ironic,” which has little relation to the English term, “ironic.” [Instructables]

OmniVision, a known quantity in the industry, beat out Apple’s current CMOS supplier, Aptina, for the contract. But the reporting one this is extremely brief and spares us virtually any details. One gets the feeling that this particular rumour was held to a lower evidentiary standard than, say, a rumour about an Apple tablet, because it’s just so plausible—most of us would be surprised if the next iPhone doesn’t boost its camera resolution to 3.2 megapixels or beyond.

Oh, and there’s the throwaway secondary rumour attached to this whole thing too: Apple has apparently also ordered a 5-megapixel CMOS sensor for “another product” set to be launched this year. Despite being able to fully ascertain that the 3.2MP sensor is meant for the “next-gen iPhone”, Digitimes’ source evidently couldn’t tell them what the other sensor is for. Hence the salt. [Digitimes via iPhoneBuzz]

There’s a rumour afoot suggesting that Canon will be ditching CCD and adopting CMOS chips for a new pro-level camcorder. Digital cameras and camcorders have never been so indistinguishable.

If the rumour is true, Canon will be adopting sub-35mm dSLR sensors (APS-C sized CMOS, or what you see in entry level dSLRs like the Rebel) into their elite camcorders. It’s not a completely new idea. The Red One has long used a CMOS chip to record 4k video, and Canon makes use of a CMOS in the $US1000ish Vixia. But with Canon choosing CMOS for a pro-level camcorder, it pretty much means that CCD (the preferred video chip format of the last several decades) is dead. (Once we saw dSLRs shooting 1080p, we knew this day wasn’t far off.) As for the mystery cam itself:

The sub-$US8,000 camcorder is said to resemble the XL-H1 (above), accepting EOS lenses and featuring a 12.1MP CMOS that can film 1080p video at 60fps/120hz—that’s MPEG4 encoded at a max rate of 56Mbps. We’re not sure how the camera will record this much data though the Red One offers CompactFlash, RAID and SSD options. There’s also word of a 12bit video RAW format that will require a $US4,000ish IO box providing SDI and USB 3 output.

And for the first time in some time, Canon’s prosumer camcorders are exciting again. [canonrumors]

DxO Labs does something quite tricky: Quantifies the quality of an image sensor’s performance, so you can look at a neat number grade for a quick gauge. Surprisingly, a Sony DSLR bests equivalent Canon models.

In their test—which uses the camera’s raw output as the basis for measuring sensor performance, evaluating colour, dynamic range and low-light shooting—Sony’s $US1100 A700 outscored Canon’s 40D and 50D by a few points, with a 66.3, falling just behind Nikon’s D300 and the Pentax K10D. Interestingly, the top scoring sensor of that size is in the Nikon D90, which has a whopping 72.6, 10 points ahead of the Canon 50D’s 62.9.

Okay, Sony’s Alpha cameras’ solid sensor performance actually isn’t that surprising, since Sony has a ton of experience making sensors—the Nikon D300 we lurrrrb and that does pretty well in DxO Labs’ tests, for instance, packs a Sony sensor. DxO Labs’ comparison tool, which has pretty much every major DSLR in its database, is actually pretty neat, if you believe their scores, anyway. [DxO Labs via CNET]

It uses four NXT microcontrollers with various Lego sensors and motors. It is very easy to program, each position uses an array element containing 3 motor positions, 3 motor speeds and an action such as grip, release or pause. I can easily teach it to pick anything up as long as it can reach it and it will fit into the end effecter.

Also worth noting is that the rig is actually only running at 35% speed – however, speeding it up reduces accuracy and causes the rig to shake a bit too much.

In any case, now that we know that industrial robots can be rebuilt using nothing more than Lego, isn’t it about time we start up our own factories to make our own gadgets?
[YouTube clip - Thanks Chris!]