Newer posts are loading.
You are at the newest post.
Click here to check if anything new just came in.

May 31 2012

Using Python for Computer Vision

Python is a tremendous asset when you're trying to classify images, track changes in scenes, search for items within images, implement augmented reality, or do the myriad other things that fall under the umbrella of Computer Vision. In this interview, Jan Erik Solem, author of the upcoming book "Programming Computer Vision with Python," describes the uses for some common operations, and choices programmers have.

Highlights from the full video interview include:

  • The value of Python in computer vision [Discussed at the 0:24 mark]
  • Searching for images within images [Discussed at the 2:13 mark]
  • Clustering or grouping images [Discussed at the 3:22 mark]
  • Constructing a 3D scene from images [Discussed at the 6:11 mark]
  • Modeling and calibrating a camera [Discussed at the 7:22 mark]

You can view the entire conversation in the following video:

OSCON 2012 — Join the world's open source pioneers, builders, and innovators July 16-20 in Portland, Oregon. Learn about open development, challenge your assumptions, and fire up your brain.

Save 20% on registration with the code RADAR


Related:



May 02 2011

Four short links: 2 May 2011

  1. Chinese Internet Cafes (Bryce Roberts) -- a good quick read. My note: people valued the same things in Internet cafes that they value in public libraries, and the uses are very similar. They pose a similar threat to the already-successful, which is why public libraries are threatened in many Western countries.
  2. SIFT -- the Scale Invariant Feature Transform library, built on OpenCV, is a method to detect distinctive, invariant image feature points, which easily can be matched between images to perform tasks such as object detection and recognition, or to compute geometrical transformations between images. The licensing seems dodgy--MIT code but lots of "this isn't a license to use the patent!" warnings in the LICENSE file. (via Joshua Schachter)
  3. The Secret Life of Libraries (Guardian) -- I like the idea of the most-stolen-books revealing something about a region; it's an aspect of data revealing truth. For a while, Terry Pratchett was the most-shoplifted author in England but newspapers rarely carried articles about him or mentioned his books (because they were genre fiction not "real" literature). (via Brian Flaherty)
  4. Sweble -- MediaWiki parser library. Until today, Wikitext had been poorly defined. There was no grammar, no defined processing rules, and no defined output like a DOM tree based on a well defined document object model. This is to say, the content of Wikipedia is stored in a format that is not an open standard. The format is defined by 5000 lines of php code (the parse function of MediaWiki). That code may be open source, but it is incomprehensible to most. That’s why there are 30+ failed attempts at writing alternative parsers. (via Dirk Riehle)

March 22 2011

Four short links: 22 March 2011

  1. EveryBlock Redesigned -- EB has been defined for a while now as "that site that makes my city's statistics useful and relevant". Now they're getting more into the user-reporting: As valuable as automated updates of crime, media mentions, and other EveryBlock news are, contributions from your fellow neighbors are significantly more meaningful and useful. While we're not removing our existing aggregation of public records and other neighborhood information (more on this in a bit), we've come to realize that human participation is essential, not only as a layer on top but as the bedrock of the site. They have a new mission: our goal is to help you make your block a better place. That's a bold goal, and quite a big change from where they were at. Will they manage any aspect of journalism, or will this become a GroupOn-ad-filled geo-portal for MSNBC? Looking forward to finding out.
  2. Typography in 8 Bits: System Fonts -- nifty rundown of fonts from the microcomputer days. I still go a bit weak-kneed at the sight of the C64 fonts. Which aspect of the system you're building will be remembered with weak knees in (gulp) thirty years' time? (via Joshua Schachter)
  3. Twitter in the Christchurch Earthquake -- analysis of the tweets around the quake, including words and retweets. (via Richard Wood)
  4. ChumbyCV -- computer vision framework for Chumby. CV on low-power ubiquitous hardware makes devices smarter and be higher-level sensors of activity and objects. (via BERG London)

October 15 2010

October 01 2010

Four short links: 1 October 2010

  1. Interview with Martin Wichary (Ajaxian) -- interview with the creator of Google's Pacman logo, the original HTML5 slide deck. One of the first popular home video game consoles was 1977's Atari VCS 2600. It was an incredibly simple piece of hardware. It didn't even have video memory - you literally had to construct pixels just moments before they were handed to the electron gun. It was designed for very specific, trivial games: two players, some bullets and a very sparse background. All the launch games looked like that. But within five years, companies figured out how to make games like Pitfall, which were much, much cooler and more sophisticated. Here's the kicker: if you were to take those games, go back in time, and show them even to the *creators* of VCS, I bet they would tell you "Naah, it's impossible to do that. The hardware we just put together won't ever be able to handle this." Likewise, if you were to take Google Maps or iPhone Web apps, take your deLorean to 1991 and show them to Tim Berners-Lee, he'd be all like "get the hell out of here." (via Russ Weakley)
  2. Liberating Lives -- The historian Tim Hitchcock, behind projects such as the Old Bailey Online and London Lives, has reflected on the impact of digitisation on our access to archives. Archives, he notes, tend to reflect the assumptions and practices of the institutions that created them. But by providing new ways into these records systems, technology can undermine the power relations that persist within their structures. Read the entire post, which has a moving description of the bureaucracy of Australia's racism and the modern-day projects built on it. (via spanishmanners on Twitter)
  3. Deblurring Images -- interesting research work reconstructing original scenes from blurred images. (via anselm on Twitter)
  4. 50 Years of Cyborgs: I Have Not the Words (Quinn Norton) -- We need language that lets us talk about the terrorism of little changes. Be they good or bad, they are terrible in aggregate. Thought-provoking essay pushing our ideas of change, future, technology, and culture until they break. (via kevinmarks on Twitter)

August 04 2010

Four short links: 4 August 2010

  1. FuXi -- Python-based, bi-directional logical reasoning system for the semantic web from the folks at the Open Knowledge Foundation. (via About Inferencing)
  2. Harness the Power of Being an Internet -- I learn by trying to build something, there's no other way I can discover the devils-in-the-details. Unfortunately that's an incredibly inefficient way to gain knowledge. I basically wander around stepping on every rake in the grass, while the A Students memorize someone else's route and carefully pick their way across the lawn without incident. My only saving graces are that every now and again I discover a better path, and faced with a completely new lawn I have an instinct for where the rakes are.
  3. Stack Overflow's Curated Folksonomy -- community-driven tag synonym system to reduce the chaos of different names for the same thing. (via Skud)
  4. Image Deblurring using Inertial Measurement Sensors (Microsoft Research) -- using Arduino to correct motion blur. (via Jon Oxer)

May 05 2010

Four short links: 5 May 2010

  1. Sketch for Processing -- an IDE for Processing based on Mozilla's Bespin.
  2. British Election Results to be Broadcast on Big Ben -- the monument is the message. Lovely integration of real-time data and architecture, an early step for urban infrastructure as display.
  3. Face.com API -- an alpha API for face recognition.
  4. Average Number of Books/Kindle -- short spreadsheet figuring out, from cited numbers. (Spoiler: the answer is 27)

April 28 2010

January 22 2010

Four short links: 22 January 2010

  1. notmuch -- commandline tagging and fast search for a mailbox, regardless of which mail client you use.
  2. Processing for Android -- pre-release versions of a Processing for Android devices. Mobile visual programming makes for interesting possibilities.
  3. Binary Body Double: Microsoft Reveals the Science Behind Project Natal for Xbox 360 -- machine learning to recognize poses and render in the game at 30fps. It's a basic real-time mocap and render.
  4. The Monetization Paradox -- interesting post by Charlie Stross about the quandry of authors. he proposed $9.99 cap on ebooks replaces the high-end $24 hardcover. Not only does it mean less royalties for the authors, it means less money for the publishers — or, more importantly, their marketing divisions. Here's a hint: if I wanted to spend my time marketing my books I'd have gone into marketing. I'm a writer. Every hour spent on marketing activities is an hour spent not writing. Ditto editing, proofreading, commissioning cover art, and so on. This is what I have publishers for.

December 08 2009

Four short links: 8 December 2009

  1. Python's Moratorium -- Python language designers have declared a moratorium on enhancement proposals (feature requests) while the world's Python programmers get used to the last batch of New And Shiny they shipped. I'm reasonably sure that the ALGOL designers went through exactly the same discussions, and I know Perl did too. So, don’t be afraid of it - don’t think that Python is evolutionarily dead - it’s not. We’re taking a stability and adoption break, a breather. We’re doing this to help users and developers, not to just be able to say “no” to every random idea sent to python-ideas, and not because we’re done. Reminds me of Perl god Jarkko Hietaniemi's signature file: "There is this special biologist word we use for 'stable'. It is 'dead'. -- Jack Cohen.
  2. This Week's Finds in Mathematical Physics -- I can't meaningfully contribute to the math, but golly them pictures are purty! (via Hacker News)
  3. x86 Assembly Encounter -- To use a construction industry metaphor, an average x86 assembler has the complexity and usefulness of a hammer, while the DSP world is using high-speed mag-rail blast-o-matic nail guns with automatic feeders and superconducting magnets. [...] I find it ridiculous that the most popular computing platform in the world does not have a decent assembler. What’s even worse, from the discussions I’ve seen on the net, people are mostly interested in how fast the assembler is (?!) rather than how much time it saves the programmer. (via Hacker News)
  4. Finding Tennis Courts in Aerial Photos -- more hacking with computer vision techniques and publicly-available data. This is going to lead to good things (and some unpleasant surprises, as that which was formerly "too hard to find" ceases to be so). (via Simon Willison)

November 27 2009

Four short links: 27 November 2009

  1. ProFORMA -- software which builds a 3D model as you rotate an object in front of your webcam. Check out the video below. (via Wired)
  2. BiwaScheme -- a Scheme interpreter written in Javascript. (via Hacker News)
  3. YMacs -- in-browser EMACS written in Javascript. Emacs, for those of you who were left in any doubt, is the only editor ever created by software engineers worth a damn (where "worth a damn" == "has possibly already achieved sentience") with the possible exception of teco.
  4. Historic Documents in Computer Science -- my eye was caught by John Backus's first FORTRAN manual, Niklaus Wirth's original Pascal paper, the BCPL reference manual (the C programming language got its name from the C in BCPL), and Eckert and Mauchly's ENIAC patent. (via Hacker News)

November 23 2009

Four short links: 23 November 2009

  1. Top E-Tailers Profiting From Scams -- Vertrue, Webloyalty, and Affinion generated more than $1.4 billion by "misleading" Web shoppers, said members of the U.S. Senate Committee on Commerce, Science and Transportation. [...] The government says the investigation shows that [the companies] "trick" consumers into entering their e-mail address just before they complete purchases at sites such as Orbitz, Priceline.com, Buy.com, 1-800 Flowers, Continental Airlines, Fandango, and Classmates.com. A Web ad, which many consumers say appears to be from the retailer, offers them cash back or coupon if they key in their e-mail address.
  2. Image Swirl (Google Labs) -- interesting image search result navigator. It's fun to play with, trying to figure out why particular sets of images are grouped together.
  3. Create Crisis (Dan Meyer) -- great call to arms for educators. It's still astonishing to me how few "learning xyz" books follow this advice. Would-be authors, take note! If there were ever an easy way to make your computer book stand out for being better than the rest, this is it!
  4. Typographic Character Coasters -- the single best argument for laser cutters evar. Send the patterns to Ponoko if you don't have a laser cutter handy.

November 16 2009

Four short links: 16 November 2009

  1. Choose Your Own Adventure -- numerical and visual analysis of the Choose Your Own Adventure novels. The distinguishing characteristic of My Kind Of People is that they appreciate the quantitative study of the commonplace. (via Bryan O'Sullivan)
  2. Tracking Droid Numbers -- uLocate, the makers of the Where app for Android, have been tracking the growth of the Droid phone using the data they get from the Android app store. (via BoyGenius Report)
  3. Fly Eyes Makes Better Robot Vision -- to make smaller flying robots, researchers would like to find a simpler way of processing motion. Inspiration has come from the lowly fly, which uses just a relative handful of neurons to maneuver with extraordinary dexterity. And for more than a decade, O’Carroll and other researchers researchers have painstakingly studied the optical flight circuits of flies, measuring their cell-by-cell activity and turning evolution’s solutions into a set of computational principles. [...] Intriguingly, the algorithm doesn’t work nearly as well if any one operation is omitted. The sum is greater than the whole, and O’Carroll and Brinkworth don’t know why. Because the parameters are in constant feedback-driven flux, it produces a cascade of non-linear equations that are difficult to untangle in retrospect, and almost impossible to predict. (via Slashdot)
  4. Meat Band Aids and Mass Production of Living Tissue -- Apligraf is a matrix of cow collagen, human fibroblasts and keratinocyte stem cells (from discarded circumcisions), that, when applied to chronic wounds (particularly nasty problems like diabetic sores), can seed healing and regeneration. This Gizmodo Q&A is informative.

Older posts are this way If this message doesn't go away, click anywhere on the page to continue loading posts.
Could not load more posts
Maybe Soup is currently being updated? I'll try again automatically in a few seconds...
Just a second, loading more posts...
You've reached the end.

Don't be the product, buy the product!

Schweinderl