VizWorld.com

In an amazing detailed post over at Impulse Adventure, they break down all the different settings of JPEG images and found one interesting little quirk in how Photoshop handles JPEG Images.  The moral of the story: Never use JPEG Quality 7.

What many people don’t know is that there is a quirk in the way that Photoshop defines its quality range. As mentioned earlier, Quality level 6 is the last point in which chroma subsampling is used. At Quality level 7 and higher, no chroma subsampling is used at all. With the amount of color information encoded now doubled, the file size would have naturally increased significantly at this level versus the previous level.

However, it is likely that Adobe decided to allocate the various quality levels with some relationship to the final compressed file size. Therefore, Adobe chose a poorer luminance and chrominance compression quality (i.e. higher level of compression) in Quality level 7 than Quality level 6!

What this means is that the image quality of Quality level 7 is actually lower than Quality level 6 (at least from luminance detail perspective).

This fact has apparently been confirmed with subjective MOS scores against various images at both quality levels.

Get all the details at their site.

via ImpulseAdventure – JPEG Quality and Quantization Tables for Digital Cameras, Photoshop. via Gizmodo

Graphics, Science image processing, photoshop

Click for Fullsize

Another day, another image optimization tool.  This one is called ‘RIOT’, and while it only runs on Windows it enables a nice side-by-side view of your images so that you can interactive tweak parameters and compare the result to the original image.

Radical Image Optimization Tool (RIOT for short) is a free image optimizer that will let you to visually adjust compression parameters while keeping minimum filesize.

It uses with a side by side (dual view) or single viewinterface to compare the original with the optimized image in real time and instantly see the resulting file size.

Could be a great tool for folks looking to maximize individual images (webmasters for example hoping to optimize bandwidth).

RIOT – Radical Image Optimization Tool.

Graphics image processing, riot, software

A clever use of augmented reality is making an appearance in a Takashimaya department story in Shinjuku.  The “Digital Cosmetic Mirror” by Shiseido uses some advanced image processing and facial recognition software to augment video of yourself with various cosmetics and mascara in real time.

All you do is sit down and let the camera scan your face. The terminal then gives you tailored recommendations. Pressing a few buttons on the touch-screen paints make-up onto your image in realtime, allowing you to see the results instantly. You also try out make-up that is currently making waves, along with printing out before and after photos with product information for you to go make the purchase of whichever colors caught your eye the most.

A powerful use of “tryvertising”, I would love to know just what impact this has had on sales.  See video of it in-action after the break.

via Augmented Reality Cosmetic Mirror in Tokyo.

Read more…

Science advertising, augmented reality, image processing, japan

Art forgery is big business, and detecting the forgeries is time-consuming and difficult.  New research from scientists published in the newest issue of PNAS shows a computerized algorithm based on vision research and sparse coding models that has great promise.

In their experiments, the researchers used the works of Flemish artist Pieter Bruegel the Elder as guinea pigs. Bruegel's art often depicts pastoral scenes, an image type that sparse coding is very good at processing, and his work spawned many imitators in his time, making him an ideal subject. First, the researchers took several known authentic Bruegel works, rendered them in grayscale, and processed them with sparse coding in tiny patches of 8×8, 12×12, and 16×16 pixels in different trials, making them into maps of math functions.

Once this was done, they processed patches from other drawings onto the authentic work, and looked at the probability of whether the new image's functions mapped onto the original. If the average probability peak for each pixel comparison was very sharp, the second image was more likely to be authentic; if the peak was more spread out or less prominent, the second image was probably a fake.

It has limitations, however.  The paintings used as reference need to be of similar subjects as the suspected forgery, and the algorithm is best at landscapes.  However, it’s a interesting piece of research that shows promise if further developed.

via Using computerized vision analysis to spot fake art.

Science art, image processing

Users of wide-format Projector screens may want to pay attention to a new announcement from Calibre UK.  At the ISE2010, they will be demonstrating their new HQView video scaler-switchers with full image processing support for seamless blending, warping, and more.

Comprising eight different compact or IU rack mountable and affordable models the HQView line-up offers the integrator a wide range of input types and configurations from simple format conversion at the start of the range through to the full warp mapping and soft-edge seamless blending of multiple projected images offered by the most sophisticated models, customers can select between operating in low-latency or highest image-quality modes, with Picture-In-Picture (PIP), Picture-and-Picture (PAP) and Picture-on-Picture (POP) output all possible.

More details are available on their site.

via Welcome to Calibre UK Ltd.

Hardware calibre uk, hqview, image processing, projector

Incogna’s Shopachu.com website implements a complex computer vision algorithm to allow users to “Find Similar Products” with nothing but an image to go on, and uses the power of NVidia’s CUDA libraries to make it happen in real-time.

Visual Guided Navigation starts by analyzing the patterns, shapes, texture and color in images to help cluster and associate similar-looking images. This process is complex and computationally intensive, with many variables that have to be taken into account, such as lighting, camera angle, background noise and photo quality. In fact, this sort of computer vision problem is still an intensively researched topic today, and Shopachu is one of the few commercial deployments to have scaled the algorithms successfully.

See a video demonstration of the system after the break.

via nTersect Blog – NVIDIA.

Read more…

Science cuda, image processing, nvidia, shopachu, Website

Google has rolled a new visualization tool for their powerful image searching system that combines technology available in some of their other products with powerful image recognition technologies to allow you to view images in a graph-layout mapped by similarity.

Image Swirl expands on technologies developed for Similar Images and Picasa Face Recognition to discern how images should be grouped together and build hierarchies out of these groups. Each thumbnail on the initial results page represents an algorithmically-determined representative group of images with similar appearance and meaning. These aren’t just the most relevant images — they are the most relevant groups of images.

Currently, it only seems to work for a fixed number of pre-selected queries (200,000 at time of writing).

via Official Google Blog: Explore images with Google Image Swirl, now in Labs. via Infosthetics

Science google, image processing

Earlier this month, Strata announced a new version of Foto3d CX with improved accuracy and several new features.  The idea of scanning physical models, reversing the traditional digital prototype to physical model pipeline, is not new and laser scanners have been around for some time.  Doing this with nothing but a run-of-the-mill digital camera intrigued me, and Strata was happy to oblige my curiosity with a version of the software to review.

Read all about it after the break.

Read more…

Graphics, Science feature, foto3d, image processing, review, software, strata

The Java Neural Network Framework “Neuroph” has just been updated with a few new routines to aid in Image Processing work, mostly image recognition.  On their website they’ve posted some examples and howto’s demonstrating what you can do.

Neural networks are one technique which can be used for image recognition. This tutorial will show you how to use multi layer perceptron neural network for image recognition. The Neuroph has built in support for image recognition, and specialised GUI tool for training image recognition neural networks. Simple image recognition library can be found in org.neuroph.contrib.imgrec package, while image recognition GUI tool is in easyNeurons application [Main Menu > Tools > Image recognition]

via Image Recognition with Neural Networks HowTo.

Science algorithm, api, image processing, java, library, neural network

It’s making a huge buzz on the internet right now and if you’re heading to SIGGRAPH ASIA then you’ll get a chance to hear alot more about it.  What is it? PhotoSketch.

We present a system that composes a realistic picture from a simple freehand sketch annotated with text labels. The composed picture is generated by seamlessly stitching several photographs in agreement with the sketch and text labels; these are found by searching the Internet. Although online image search generates many inappropriate results, our system is able to automatically select suitable photographs to generate a high quality composition, using a filtering scheme to exclude undesirable images. We also provide a novel image blending algorithm to allow seamless image composition.
Each blending result is given a numeric score, allowing us to find an optimal combination of discovered images. Experimental results show the method is very successful; we also evaluate our system using the results from two user studies.

Cliff notes version: Draw a basic sketch of what you want and label the components (as shown above), and the system automatically extracts the necessary components from a huge database and composites them together.  The research consists of 2 main parts:  A new blending algorithm to make the images fit in their new environments much better, and a new filtering and classification system for images.

You can read the full PDF of their work here (with many more examples than the sailboat/kiss floating around everywhere), and see a great demonstration video after the break.

Read more…

Science algorithm, image processing, siggraph