• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Combating Camera Blur

John_Keck

Limp Gawd
Joined
May 3, 2010
Messages
379
Four researchers at Microsoft have developed a hardware sensor attachment to measure hand wobble and shaking to help analyze and correct blurry pictures.

"Our method is completely automatic, handles per-pixel, spatially-varying blur, and out-performs the current leading image-based methods"
 
Is that an aftermarket bra or a true mazdaspeed one? I heard bad things about the aftermarket ones.:confused:
 
Personally if they are gonna stick a car-bra on an RX-8 then the blurrier the picture the better.

If you're gonna show off an anti-blur tech then take clean up some pics of hot celebs topless dammit!
 
Hope it better works better than the image stabilization in my Canon powershot. The slightest shake and I get a crappy pic.
 
I'm a professional photographer, I want this INCLUDED in cameras........
 
I thought I was still browsing craigslist looking at cars when I visited the front page.

Oh, and remove that bra. (*insert that what he said comments here*)
 
That's not camera blur, that's what a rotary looks like during catastrophic engine failure.
 
The guy who wrote the paper has some amazing projects. Just wasted about an hour on the MS Research site.
 
who needs "a combination of inexpensive gyroscopes and accelerometers in an energy optimization framework to estimate a blur function from the camera’s acceleration and angular velocity during an exposure." Is it that hard to use a tripod?
 
from article said:
To the best of our knowledge, this is the first work that uses 6 DOF inertial sensors for dense, per-pixel spatially-varying image deblurring and the first work to gather dense ground-truth measurements for camera-shake blur.

Just wait 10 years! Someone will come out of the woodwork and present a patent they sat on for 20 years until Microsoft make millions! lol

Anyways, I would love for this technology to come into fruition. I don't make the most perfect photos and this might help me and lots of others big time.
 
who needs "a combination of inexpensive gyroscopes and accelerometers in an energy optimization framework to estimate a blur function from the camera’s acceleration and angular velocity during an exposure." Is it that hard to use a tripod?

Considering 99% of the pictures taken are from the hands and no one is going to lug a tri-pod around with them to take pictures on the go ... yes it is "that" hard to use a trip pod.

I have nerve damage in my left hand and something forming in my right. Both my hands shake so when ever I take pictures at parties or whatever my pictures generally suck even with the current "anti-blur" technology in these camera's, this would be epic for me!
 
who needs "a combination of inexpensive gyroscopes and accelerometers in an energy optimization framework to estimate a blur function from the camera’s acceleration and angular velocity during an exposure." Is it that hard to use a tripod?

Incredibly simple to use a tripod, but you can't always have one with you. ;)

Would be nice to have this built into the camera image sensor, so no more need for image stabilization in the optics. You could also turn this on or off when you process the images (I would assume?)
 
Hope it better works better than the image stabilization in my Canon powershot. The slightest shake and I get a crappy pic.

Strange, I find the image stabalization in both my Canon cameras to be very useful. Guess I have a steady hand. On my SX-10, I've taken hand held pitures at the full 20x zoom that have turned out very sharp. I took a few pictures of a lion in a zoo in Australia, and you can see flies sitting on the lion's nose.

My biggest problem is with motion blur in low light (due to objects in the picture moving during exposure).
This technology still does nothing to help with motion blur.
 
Looks better, but still a bit blurred to be though. For the best pictures, I suggest a DSLR. I know... they're expensive...
 
who needs "a combination of inexpensive gyroscopes and accelerometers in an energy optimization framework to estimate a blur function from the camera’s acceleration and angular velocity during an exposure." Is it that hard to use a tripod?

Most typical user don't have that, or carries one around with them when they go travelling.

This can be very useful to many people
 
By 'inexpensive', are they referring to a regular camera, or profesional ones? A professional camera would likely have a tripod nearby anyway. But if it's for the regular-carry-everywhere-vacation camera, this is a very big deal.
 
Best option for good pictures is to pay someone else to take them. If their wobbly you don't pay.
 
I'm so tired of the need for VR lenses for my Nikon 300. This would be a photographers god send. Could mean much faster shutter speeds and smaller AP settings. This is a low light wish list for ANY photographer. I want... and want it now.
 
Would also be useful not just for stills, but for removing blur from video, especially when combined with shake stabilization
 
I'm a professional photographer, I want this INCLUDED in cameras........

The image stabilization that is built into my Pentax K10D body works VERY VERY well.

I don't have a mono or a tripod and even if I don't hold the camera super still I can get good pics at close to 0.5s exposure time.

I tend to think that it is a lot better then Nikon's solution which is to have the image stabilization built into the lenses.

Why even have a camera that you have to buy special lenses for in order to have image stabilization?
 
So how is this different than what is currently used in vibration reduction in SLR lenses, like for instance my Nikon 70-200 f/2.8?



Is the difference that the VR systems in the SLR lenses is mechanical, and this is a chip based digital solution?
 
I don't think the input is any different than what is used for in camera body image stabilization.

however if you look at the high res results for their deblurring, the results are ass-tastic. I'd argue worse than being blurred, and worse than using deblurring software with a hunman tweaking it till it looks least bad.

The reduced image size is hiding a lot of suck.
 
who needs "a combination of inexpensive gyroscopes and accelerometers in an energy optimization framework to estimate a blur function from the camera’s acceleration and angular velocity during an exposure." Is it that hard to use a tripod?
That's exactly what everyone wants to do with their pocket sized point and shoots....carry a tripod around :rolleyes:
 
If you really need a bra, get the 3M clear bra.

I miss my '8. :(
How well do those work? My Celica has so many chips and dings on the hood (from rocks on the freeway...) that it's not even funny :(

But on-topic, it would be nice to see more image stabilization features in cameras in the future. I haven't had the time nor funds to go get a VR lens for my D200 yet, and I have quite possibly the most unsteady hands ever.
 
They need de-blurring for camera phones since it's often a pain to hold them steady.
 
That's exactly what everyone wants to do with their pocket sized point and shoots....carry a tripod around :rolleyes:

lets see you fit the camera mentioned in your pocket, dumbass.

teaser.png
 
lets see you fit the camera mentioned in your pocket, dumbass.

teaser.png
Wow, you're a real piece of work. It's a prototype that will more than likely make it's way into consumer point and shoots. If anything you're doing a great job of making yourself look like a complete idiot in this thread.
 
I still don't see what's new here.

We've had this kind of technology in SLR's for years.

What does this do, that is an improvement over existing solutionslike Nikon's VR, Canon's IS ?

Is it significantly cheaper, smaller or more effective than previous solutions?

Otherwise reading this thread is sort of like me posting something about a genius new prototype video card that - believe it or not - can actually accelerate 3d!!!
 
Zarathustra[H];1036013804 said:
I still don't see what's new here.

We've had this kind of technology in SLR's for years.

What does this do, that is an improvement over existing solutionslike Nikon's VR, Canon's IS ?

Is it significantly cheaper, smaller or more effective than previous solutions?

Otherwise reading this thread is sort of like me posting something about a genius new prototype video card that - believe it or not - can actually accelerate 3d!!!
They explain the introduction in the paper, the link is only of the abstract. I hate to be "that guy" and tell you to read the link but I didn't read it all the way through yet
 
They explain the introduction in the paper, the link is only of the abstract. I hate to be "that guy" and tell you to read the link but I didn't read it all the way through yet

Doh, my bad. I didn't realize there were links to the full thing on the bottom.
 
Zarathustra[H];1036014045 said:
Doh, my bad. I didn't realize there were links to the full thing on the bottom.
I didn't either until your comments got me wondering what the difference was.
 
Back
Top