Features and Examples
Blurity is the most capable blur-removal tool on the market today.
With Blurity, you too can remove the blur from your blurry photos. All you need is a computer, a fuzzy photo, and the desire to see it sharp. You can leave Blurity with its default settings for easy use, or you can dive into the settings if you're a power user and want to squeeze the best deblurring possible into your photo.
- Example: Forensic text recovery
- Example: Out-of-focus goalie
- Example: Complex motion blur
- Example: From a real customer
- Comparison versus the competition
Here are some real examples of Blurity in action. All of the photos were taken with real cameras and show real blurs.
You can replicate these results for yourself if you install Blurity on your PC.
Forensic text recovery
Let's say you took a photo of some text, perhaps as a note to yourself, but the light was bad. Or maybe you're a police detective, and this was a critical piece of evidence you found. Regardless, you're dealing with a blurry photo.
Maybe it looks something like this photo of a TI-89 graphing calculator screen, taken with a Motorola Droid X phone camera.
Can you read what the screen says? Unless your eyes are a lot better than normal, probably not. What if you really needed those numbers that were displayed on the screen? In the past, you would have been out of luck. These days, you have Blurity.
Now let's load it into Blurity. The blur size looks pretty extreme, so we'll choose a blur size of 37 pixels to ensure that we get it all, a 450 pixel sample box box size to ensure we have enough data to build the model, a solver filtering level of 12 to reduce noise in the blur model, and 10 solver iterations to keep the solver from falling into a false solution too early.
The area around the yellow and purple keys on the calculator was chosen for the sample box, since that spot offers both strong edges and fine detail without being over- or under-exposed.
The image in Blurity looks like this:
Now that Blurity has enhanced the image and removed the motion blur, it's easy to read the text on the screen. You can see that the first expression was 1680/1050, the second was .5/.015, and the third was 1/.015.
Yes, the image still has some artifacts: this is normal and expected. The causes are a combination of sensor noise, overexposure, underexposure, non-uniform blur around the image, and the assumptions made by the Fast Fourier Transform (FFT).
But you don't have to worry about that.
Try it yourself! Download Blurity and get started deblurring!
Out of focus goalie
Even with all our modern whiz-bang technology, cameras still don't focus correctly all of the time. Sports can be especially tricky for focus, since there are lots of things moving and the light is usually pretty dim.
Here's an example of a hockey goalie that's a bit out of focus and very slightly motion blurred. Unlike our previous example, this image was captured with a high-end camera and lens: a Canon 7D with a 70-200mm f/2.8 IS lens.
It would be a pretty nice shot of the goalie in his stance, but the slight blur makes it unusable. Let's see what we can do with Blurity.
The default settings in Blurity work great on this image.
The goalie's chest was chosen for the sample box location due to the detail on the jersey crest and the hard edges offered by the white glove against the blue jersey.
The net would have been a bad choice for the sample box location, since it consists mainly of small lines, which can confuse the blur modeling algorithm.
The end result. Blurity has removed most of the focus blur as well as the slight motion blur.
The improvements can be subtle but noticeable sometimes. The improvement is most apparent if you look at the white concentric circles on the goalie's jersey's crest. On the deblurred image, there is clear separation between the circles around their entire circumference.
Could you get a similar result in Photoshop for this image? Yes, Smart Sharpen does a decent job with slightly out-of-focus images. However, the great advantage for Blurity here is that the blur removal/sharpening occurred without any settings adjustments. In Photoshop, you would have had to tweak things to get a similar result.
Want to give it a try? Download Blurity and get started deblurring!
Complex motion blur
Sometimes you get an image that's just incredibly blurry, like this one. It was taken with a Canon A1100 point-and-shoot.
Can you read the brand name on the mug? How about the date on the watch's little date window? Believe it or not, you'll be able to read it clearly after Blurity!
Now let's bring the image into Blurity.
To be honest, it took a bit of trial and error to find the best location for the red sample box. In this case, the best location was the watch face. Switch to Basic mode, set the Blur Level to Normal, and place the sample box on the watch face. Boom: deblurred.
Not only did Blurity correctly and successfully fix the blur, it did it very quickly. On our test computer, processing this image took 11 seconds. In other words, Blurity is 32 times faster than SmartDeblur (which took 300 seconds), and unlike SmartDeblur, Blurity actually removed the blur.
Blurity did a great job recovering the detail!
It's easy to read the brand name on the mug ("Porsche") and the date in the little white window on the watch ("3"). Even the small tic marks on the sub-dials on the watch face are clear if you zoom in all the way!
Yes, there are artifacts, most noticeably "ringing," but that's to be expected with such a large blur.
Enahance! Download Blurity and get started deblurring!
From a real customer
A customer named John emailed us recently. In an excited tone, he described how he was using Blurity to successfully deblur many of his photos. He kindly allowed us to share this example of his.
From this blurry version, we can see what appears to be motion blur in a generally vertical direction. We can tell the direction because vertical lines are sharp, while horizontal lines are blurred. Thus, when we process the image with Blurity, we'll be looking for a computed blur model that looks somewhat like a vertical line.
Although the blur in this image was clearly relatively small, it took a few tries to find the optimal placement for the red sample box. Unlike many images, the best location ended up being centered on the rocks. It worked in this situation becuase of the relatively high amount of texture on the rocks. Basic-Normal parameters were sufficient.
It's also worth mentioning that John applied a smart transformation to the image before sending it to us (and before deblurring it himself): he resized what had been a very large image to one substantially smaller. By doing that, he substantially increased his odds of success. By rescaling the image to smaller dimensions, the he increased the signal-to-noise ratio of the image, which made life easier for Blurity's deblurring algorithm.
Blurity worked quite well. The level of detail is substantially improved in the deblurred image. Moreover, the blur model appears to have been relatively accurate over most of the image, as there are just a few very minor visual artifacts present.
As expected, the blur turned out to be represented by a relatively vertical (though not precisely vertical) "point spread function" (PSF) about 6 pixels tall and 3 pixels tall, with a bit of a leftward curve. This is visible in the "computed blur model" display within the Blurity interface. Another way to think about it is that the PSF traces the movement of the camera during the exposure. Blurity attempts to figure out how the camera moved, and once it has a model of the movement, it attempts to undo that movement.
This was a successful deblurring.
Unblur your photos, too: Download Blurity!