Re: multitouch + audio feedback linux dev - XO-2?
1) Are there any existing hooks/systems for Linux for multi-touch? That's the only proper way you can get a virtual keyboard to work for a double-touchscreen clamshell device (the feasibility of which is not sold to me because of the power consumption of running a 2nd screen vs a keyboard, and mostly mostly mostly the lack of haptic feedback from a virtual keyboard). Peter Hutterer has been working on Multi-Pointer X for several years. It is getting pretty usable, judging by the Youtube videos. http://wearables.unisa.edu.au/mpx/ http://www.youtube.com/watch?v=olWjnfBoY8E In fact, you could pretty much do multitouch systems with just a camera (the XO's would be good enough, the machine would just need a CPU with stronger horsepower to do the image processing :P) and any surface where discrete image sections can be formed for img recog aid, like using shadows from hands or maybe a color-reactive transparent surface where if pressure or contact is applied, a particular color shade (like a chroma key) will appear to aid img rec. You can use the pygame camera module to do multitouch on an XO, though it is far from usable for anything but drawing. http://www.youtube.com/watch?v=HIDdxY3L5V8 Nirav ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [Sugar-devel] Wacom Bamboo with XO?
Chris, We're working on switching from gstreamer to the pygame camera module. Gstreamer in 8.2 had some changes that broke the pipeline Colors was using. While it would have been possible to change the pipeline, switching to pygame will be easier/better for both the videopaint and photo snap parts. This will likely be done before the end of the month. Nirav On Mon, Dec 8, 2008 at 8:02 AM, Chris Marshall [EMAIL PROTECTED]wrote: BTW, Colors! version 11 did not appear to have the photo snap capability. Was that removed? Cheers, Chris ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Panorama activity
I applied for project hosting a few weeks ago but haven't gotten a reply yet. For now I'm just using my own repo. I'm starting up an Activity development team at my university, and our first projects are going to be getting Panorama and Bridge to a deployable state. Our first meeting is going to be this week, and there should be a more stable release of Panorama over Thanksgiving weekend. Nirav On Tue, Nov 4, 2008 at 2:57 PM, Samuel Klein [EMAIL PROTECTED] wrote: this is still one of the cooler projects people have taken on! Jeff K, have you gotten to try it out? Nirav, any thoughts of another release? Else you (or someone who wants to maintain it) should set up hosting for the code... SJ On Sun, Oct 12, 2008 at 2:15 AM, Nirav Patel [EMAIL PROTECTED] wrote: That sounds like a much better idea. I'll try to set up something like that on my webserver, though I'll have to move elsewhere if the activity gains any kind of popularity. It still leaves the problem of the images uploading to my personal Flickr account though. I think having a Panorama Collection or Set on the OLPC Flickr account would be nice. Also, I uploaded the pre-alpha bundle: http://eclecti.cc/files/Panorama-1.xo Nirav On Sun, Oct 12, 2008 at 1:39 AM, Brian Jordan [EMAIL PROTECTED] wrote: Maybe one could set up a one-way upload script for this on dev (or some appropriate community-developed-web-services-server-of-the-future)? This script, then, could handle the uploading via flickr API securely. Regards Brian ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Hosting request: Panorama
1. Project name :Panorama 2. Existing website, if any : http://wiki.laptop.org/go/Panorama 3. One-line description : A panorama stitching activity 4. Longer description : An auto panorama stitcher that can upload the : captured panoramas to a central website of some kind. 5. URLs of similar projects : http://wiki.laptop.org/go/Panorama_camera_activity 6. Committer list Please list the maintainer (lead developer) as the first entry. Only list developers who need to be given accounts so that they can commit to your project's code repository, or push their own. There is no need to list non-committer developers. Username Full name SSH2 key URLE-mail - -- #1 nrp Nirav Patel http://n0r.org/id_rsa.pub [EMAIL PROTECTED] #2 #3 ... If any developers don't have their SSH2 keys on the web, please attach them to the application e-mail. 7. Preferred development model [X] Central tree. Every developer can push his changes directly to the project's git tree. This is the standard model that will be familiar to CVS and Subversion users, and that tends to work well for most projects. [ ] Maintainer-owned tree. Every developer creates his own git tree, or multiple git trees. He periodically asks the maintainer to look at one or more of these trees, and merge changes into the maintainer-owned, main tree. This is the model used by the Linux kernel, and is well-suited to projects wishing to maintain a tighter control on code entering the main tree. If you choose the maintainer-owned tree model, but wish to set up some shared trees where all of your project's committers can commit directly, as might be the case with a discussion tree, or a tree for an individual feature, you may send us such a request by e-mail, and we will set up the tree for you. 8. Set up a project mailing list: [ ] Yes, named after our project name [ ] Yes, named __ [X] No When your project is just getting off the ground, we suggest you eschew a separate mailing list and instead keep discussion about your project on the main OLPC development list. This will give you more input and potentially attract more developers to your project; when the volume of messages related to your project reaches some critical mass, we can trivially create a separate mailing list for you. If you need multiple lists, let us know. We discourage having many mailing lists for smaller projects, as this tends to stunt the growth of your project community. You can always add more lists later. 9. Commit notifications [ ] Notification of commits to the main tree should be e-mailed to the list we chose to create above [ ] A separate mailing list, projectname-git, should be created for commit notifications [X] No commit notifications, please 10. Shell accounts As a general rule, we don't provide shell accounts to developers unless there's a demonstrated need. If you have one, please explain here, and list the usernames of the committers above needing shell access. 11. Translation [X] Set up the laptop.org Pootle server to allow translation commits to be made [ ] Translation arrangements have already been made at ___ 12. Notes/comments: ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Panorama activity
That sounds like a much better idea. I'll try to set up something like that on my webserver, though I'll have to move elsewhere if the activity gains any kind of popularity. It still leaves the problem of the images uploading to my personal Flickr account though. I think having a Panorama Collection or Set on the OLPC Flickr account would be nice. Also, I uploaded the pre-alpha bundle: http://eclecti.cc/files/Panorama-1.xo Nirav On Sun, Oct 12, 2008 at 1:39 AM, Brian Jordan [EMAIL PROTECTED] wrote: Maybe one could set up a one-way upload script for this on dev (or some appropriate community-developed-web-services-server-of-the-future)? This script, then, could handle the uploading via flickr API securely. Regards Brian ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Panorama activity
My roomate and I wrote an auto-stitching Panorama Activity during the Yahoo Hack Day at CMU yesterday (ended up winning the Hack for Good award). I'll be posting it in the wiki at some point today. It needs a whole lot of polishing, since it was written in 24 hours. I ran into the same problem that Ben did with autowhitebalance and autoexposure. I looked through the ov7670 driver, and it seems changing V4L2_CID_BRIGHTNESS should flip the auto exposure control (AEC) bit in one of the registers. I added support for changing brightness to the camera module in Pygame, but either disabling AEC doesn't work, or it just makes auto gain control and auto white balance work even harder to change the image. Other than that, the stitching seems to work ok. I'm going to pyramidize it at some point to make it faster, and do something to improve the accuracy. The slowness of stitching doesn't effect the user experience much though, because it only stitches after the user is done taking the pictures. The Activity we wrote uses the Flickr API for an upload to Flickr button. I feel that this is the most important part. There is support for saving the stitched images as Journal entries, but it would be wonderful if a kid anywhere in the world could see panoramas taken by other kids elsewhere. Flickr is far from ideal for this. Beyond not really being designed for viewing panoramas, it is currently set up to just upload every picture to my personal Flickr account. My API key, secret key, and an authentication token are prerecorded in the activity, so anyone could potentially use those values to edit existing photos and upload whatever to my Flickr account: http://flickr.com/photos/nrpatel/ It seems Gigapan doesn't have an upload API. gigapan.org works in Browse on my os767 XO, but panning around an image is pretty slow. Anyone have ideas on this? A better way to Flickr, ideas for designing a new site, getting Gigapan to add an upload API, etc? The other issue we ran into is integrating OLPCGames with pyGTK stuff. We used OLPCGames to take advantage of capturing the images as SDL surfaces, but we also wanted pyGTK to pop up a Dialog box when the user hits Save, asking for a Title, Tags, and a Description of the panorama. We hacked together something that does pop up the box and saves correctly, but fails to ever destroy the box. It then just sits with the box open, preventing the user from using or exiting the Activity. Is it possible to use pyGTK beyond the toolbox while still using OLPCGames? Nirav On Fri, Sep 5, 2008 at 6:41 PM, Jeff Keller [EMAIL PROTECTED] wrote: Yeah -- I sent my loaner unit off to Burning Man, but I should get it back soon, and I'd be delighted to talk about stitchers. Randy Sargent and I wrote our own for GigaPan last summer because nothing else was up to the task. It relies on a model of the GigaPan device's behavior to do rough placement, but it can auto-stitch 1000+ images with 10 pixels rms displacement error, and it should be open-sourced one of these days. Of course, it uses much more storage, RAM, and CPU than an XO has, but nothing inappropriate to what it does, and there are some nice improvements in the works. Oh, and my office is down the block from OLPC's. --Jeff On Fri, Sep 5, 2008 at 12:38 PM, Samuel Klein [EMAIL PROTECTED] wrote: Speaking of which, Jeff Keller has been trying to get OLPC folks to learn more about using his gigapan group's cameras (and perhaps to borrow one to bring them into a country for weeks) for a while. Jeff, Ben Schwartz and Nirav are really interested in large-scale image splicing... SJ --- 617 529 4266 On Fri, Sep 5, 2008 at 2:08 AM, Brian Jordan [EMAIL PROTECTED] wrote: *bump* http://wiki.laptop.org/go/Panorama_camera_activity (code? Nirav is interested in doing something similar!) Brian On Wed, Jun 20, 2007 at 11:56 PM, Benjamin M. Schwartz [EMAIL PROTECTED] wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 The XO happens to be perfect for shooting stitched panoramic photographs, due to the swivel design. I tested it out in the OLPCHQ lobby. Then, I wrote a simple panorama stitcher in 50 lines of Python. It runs in 3.4 seconds on my Core Duo, producing this output: http://people.fas.harvard.edu/~bmschwar/lobby_ugly.jpg The results aren't too bad. I also tried stitching this scene with Hugin, the most powerful panorama stitcher I know of. Hugin required significant user intervention and half an hour of computing time, producing this output: http://people.fas.harvard.edu/~bmschwar/lobby_pretty.jpg This scene is unusually difficult because of the huge indoor-outdoor contrast. Given this positive result, I would like to work on a panorama-making activity, possibly inside Capture. I know that at age 10, I loved making panoramas out of photographs. Panoramas provide an immersive way for children to communicate their environments to each other and to the
Re: Wellington test team vs Activities - 2008-09-13
Thanks for the bug report on Bridge. I just released Bridge-2 that fixes those bugs and a few others. Nirav 2008/9/13 Grant Patterson [EMAIL PROTECTED]: Grant's notes: Geoquiz-3 - Too easy to mistakenly select the wrong answer when using the left and right controllers (as suggested by the instructions). This problem is caused by Geoquiz not distinguishing between the left and right controllers. - Lesotho, Swaziland and Gambia aren't visible on the Africa map, so when these questions arise, the user can only guess. Swaziland and Gambia are blacked out on the map, while Lesotho appears as part of South Africa. - Completing the quiz (either Africa or South America) causes Geoquiz to crash! - The questions are always in the same order until the laptop is rebooted. - Got a segmentation fault during the normal usage of the activity! Wasn't reproducable. Stopwatch-1 - Worked well Bridge-1 - Seems a little harsh that girders that are deleted are still counted towards the total cost of the bridge, especially since they can't be moved once in place. - Some instructions would be helpful. (e.g. press space to start train) - Pressing the 'b', 'j', or 'p' keys crashes the activity! - Otherwise a great activity. Grant 2008/9/13 Tabitha Roder [EMAIL PROTECTED] Tabitha's notes: XO Build 8.2-759 Mini vMac-1 - ctrl F for full screen is great! played with excalibur, note pad, calculator, puzzle, all good, but it is hard to shut down Mini vMac - seems you have to stop in the frame? Bridge-1 good Stopwatch-1 good Poll-17 - create own poll good, voted good, lesson plans display as XML not so good, bug logged #8456 Schoolsplay-1 - loads, font size is too big in the frames and in the pyramid, cant really play as cant read :-( bug logged #8458 Audacity-1 - fails to launch, got logs, bug logged #8453 XaoS-1 - good Moon-4 good Colors-4 - still not able to stop activity, same as v4 - see original bug #8335 ProducePuzzle-0 - good Ok, now onto the big test - GCompris-7, get ready this is long GCompris - Various puzzles *assemble - good *tangram - good *build - good *superbrain - good *hanoi - good *sliding blocks - good *sudoku - good *fifteen game - good Gcompris - Mathematic games - geometry *freedraw - good *redraw - could not find or load the file skins/gartoon/redraw.svg this activity is incomplete exit it and report the problem to the authors ([EMAIL PROTECTED]) *mirror - could not find or load the file skins/gartoon/redraw_symmetry.svg this activity is incomplete exit it and report the problem to the authors ([EMAIL PROTECTED]) Gcompris - Mathematic games - calculations *math memory **multiply - good **add - good **subtract - good *algebra **add and subtract tested, took a while to work out that you have to push check answer or it thinks you havent entered an answer *number munchers **multiple - good **prime - good *math memory **bit confusing that questions with the same answer dont match in the matching game - e.g. 4 + 4 is not equal to 3 + 5, they are both = 8 only *balance the scales - good *target - good *series of correct operations - good Gcompris - Mathematic games - numeration *pairs of dice - good *count the items - couldn't find or load the file enumerate/food/banana.png this activity is incomplete - bruno *magician hat - good *feed tux the fish - good *money - couldnt find money/euro/p5e.png - bruno *helicopter - good *draw with numbers - good GCompris - amusement *tuxpaint - cannot find tux paint install it to use this activity *football - good *hexagon - good *word processor - good, we all like the themes *drawing / animation - works for draw not animation - save image exits GCompris so cannot make animation *chat - sorry couldnt test this today GCompris - reading *click on letter - good *reading practice - good *horizontal - words start above the box not in it *vertical - words start above the box not in it *missing letter - good *image name drag and drop - good GCompris - discover the computer *keyboard **numbers with dice - good **throw ball to tux - good **simple letters - good **falling words - good *mouse **control the hose pipe - good **click the mouse - good **move the mouse - good **click on me - good **click and draw - good GCompris - strategy games *practice chess - error: the external program gnuchess is mandatory to play chess in gcompris first install it and check it is in /usr/games/gnome-gnuchess *connect 4 - good *bar game - good - the prerequisite of a brain is cute *oware - good GCompris - experiental games *parachutist - good *canal lock - good *water cycle - good *electricity - cannot find the gnucap electric simulator you can download and install it from http://geda.seul.org/tools/gnucap/ to be detected it must be installed in /usr/bin/gnucap or /usr/local/bin/gnucap you can still use this activity to draw schematics without computer
Re: [sugar] [Activities] Camera raw mode (WAS: Panorama activity)
It supports Bayer RGGB8, which should be basically unprocessed data off of the CMOS. I can't imagine it being useful though. It would require unBayerizing it to display, even though the XO's display is also Bayer (I assume pixel offsets and swizzling would mess it up, even if the colors are in the same order). You're better off letting the camera do the heavy processing and just get packed RGB back. Nirav On Sat, Sep 6, 2008 at 10:15 AM, Christoph Derndorfer [EMAIL PROTECTED] wrote: Nirav, IIRC the camera also supports a RAW mode, right? Is this another feature that we could integrate into an advanced camera activity that also includes the features mentioned below? Thanks, Christoph On Fri, Sep 5, 2008 at 4:26 PM, Nirav Patel [EMAIL PROTECTED] wrote: I don't know about integrating it into Record, but the other issues are fairly easy to resolve. I wrote a Python module over the summer to interface v4l2 cameras that returns Pygame surfaces. Capturing images is near instant (~20ms for 640x480 RGB). Autogain and autowhitebalance can be disabled using v4l2 camera controls. There is support in my module for the hflip and vflip controls, but adding additional controls is trivial. Alternately, we could just use a little utility in C to access any v4l2 camera control. Since the module I wrote is for Pygame and is eventually going to be cross platform, adding v4l2 specific stuff to it is probably not the greatest idea. It may just be better to skip sticking it in Record and make a standalone activity, since Record seems to have plenty of issues on its own without a big new feature. There wouldn't be much code duplication if Panorama doesn't use gstreamer anyway. Nirav On Fri, Sep 5, 2008 at 9:26 AM, Benjamin M. Schwartz [EMAIL PROTECTED] wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Christoph Derndorfer wrote: | Sweet, I totally missed that! | | Is anyone actively working on integrating that functionality into 'record' | or making it available as a seperate activity? | | Christoph | See http://lists.laptop.org/pipermail/sugar/2008-February/004307.html The Panorama Activity is snot quite barely functional. It might be best to roll this functionality into Record, and Erik Blankinship has expressed some interest in that. However, Record's UI is very unusual, and I am not about to attempt integration with it myself. Also, as you can see in the examples, this process is only likely to work well once the camera's automatic white balance and gain control are deactivated. The only way I know how to do that is to run in Bayer mode. Bayer mode was introduced in a recent gstreamer, but gstreamer was recently downgraded, which leaves me without a known reliable way to access Bayer mode. - --Ben P.S. For Bayer mode info, see http://lists.laptop.org/pipermail/devel/2008-February/011029.html -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAkjBM3oACgkQUJT6e6HFtqRowgCfUlngIFr+Gl3jxKRYZAXBNl/x 2hEAnAnoDrrvcd+vIHO68aJthULDKAQC =cvG0 -END PGP SIGNATURE- -- Christoph Derndorfer co-editor, olpcnews url: www.olpcnews.com e-mail: [EMAIL PROTECTED] ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [sugar] [Activities] Panorama activity
I don't know about integrating it into Record, but the other issues are fairly easy to resolve. I wrote a Python module over the summer to interface v4l2 cameras that returns Pygame surfaces. Capturing images is near instant (~20ms for 640x480 RGB). Autogain and autowhitebalance can be disabled using v4l2 camera controls. There is support in my module for the hflip and vflip controls, but adding additional controls is trivial. Alternately, we could just use a little utility in C to access any v4l2 camera control. Since the module I wrote is for Pygame and is eventually going to be cross platform, adding v4l2 specific stuff to it is probably not the greatest idea. It may just be better to skip sticking it in Record and make a standalone activity, since Record seems to have plenty of issues on its own without a big new feature. There wouldn't be much code duplication if Panorama doesn't use gstreamer anyway. Nirav On Fri, Sep 5, 2008 at 9:26 AM, Benjamin M. Schwartz [EMAIL PROTECTED] wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Christoph Derndorfer wrote: | Sweet, I totally missed that! | | Is anyone actively working on integrating that functionality into 'record' | or making it available as a seperate activity? | | Christoph | See http://lists.laptop.org/pipermail/sugar/2008-February/004307.html The Panorama Activity is snot quite barely functional. It might be best to roll this functionality into Record, and Erik Blankinship has expressed some interest in that. However, Record's UI is very unusual, and I am not about to attempt integration with it myself. Also, as you can see in the examples, this process is only likely to work well once the camera's automatic white balance and gain control are deactivated. The only way I know how to do that is to run in Bayer mode. Bayer mode was introduced in a recent gstreamer, but gstreamer was recently downgraded, which leaves me without a known reliable way to access Bayer mode. - --Ben P.S. For Bayer mode info, see http://lists.laptop.org/pipermail/devel/2008-February/011029.html -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iEYEARECAAYFAkjBM3oACgkQUJT6e6HFtqRowgCfUlngIFr+Gl3jxKRYZAXBNl/x 2hEAnAnoDrrvcd+vIHO68aJthULDKAQC =cvG0 -END PGP SIGNATURE- ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Malfunctioning touchpad
I've had the same issue with my G1G1 XO. It sounds like it could be http://dev.laptop.org/ticket/5575 My problem mostly went away by itself after I took apart the XO and put it back together. Nirav 2008/6/30 pradosh Kharel [EMAIL PROTECTED]: I am troubleshooting hardware problems for the next week for Nepal's pilot schools. I will be asking questions to this list and sending my experiences. I had a laptop with a touchpad that only detected motion along the y axis. Swapping the card with an older version of the xo resulted in the touchpad not functioning at all and some keys not working. When I replaced the Keyboard of the malfunctioning XO with another one, both the keyboard and the touchpad worked fine. ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
GSoC Status Report: Vision Processing
As you may know, OLPC got GSoC students again this summer. I am one of them, and my project is Vision Processing. That is, a library to use the webcam for more than capturing images. I am implementing this by adding v4l2 and computer vision functions to pygame. My code is available at http://git.n0r.org/?p=pygame-nrp;a=summary and is currently pygame 1.8.1 with the addition of a camera module that supports v4l2 cameras that use MMAP and have pixelformats of RGB24, RGB444, YUYV, SBGGR8, or YUV420. Basic usage is as follows: import pygame from pygame import camera cam = camera.Camera(/dev/video0, (640, 480), RGB) # the third argument can be YUV or HSV too. cam.start() frame = cam.get_image() # the frame returned is a 24bit pygame Surface You can also do fun stuff like: http://eclecti.cc/bytes/living-pointillism-a-pygame-webcam-script or more practical stuff like having it track the centroid of a specific hue (green in this case): http://eclecti.cc/files/centroid.py My plans are to add functions like finding the largest connected component, optical flow, and other things useful for computer vision. Currently, performace is pretty poor on the XO; a combination of the Geode being slow and having to convert from 24bit to 16bit surfaces to display any captured frames. The XO is fast enough to capture and blit a 320x240 RGB frame at 30fps, but not at 640x480 or a frame being converted to HSV. I'm not sure how or if I'm going to be able to overcome those performance problems. I'd appreciate any comments, suggestions, or reality checks on improving performance or anything else, or any requests for vision functions to add. Also, I only have the camera in the XO, vivi, and a poorly supported USB webcam, so if anyone could test it on other webcams, that would be great. Thanks, Nirav Patel ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Face Detection and other Vision Processing (and some sample code)
I've been playing with vision processing stuff for the XO, mostly using OpenCV. I have some sample code up at http://eclecti.cc/code/face-detection-on-the-olpc-xo that does face detection surprisingly quickly. It seems that the bottleneck isn't even the OpenCV Haar Cascade algorithm, but the time it takes to initialize the v4l2 drivers. The code I wrote is just basic proof of concept stuff, but there are some amazing possibilities involved in vision processing, like recognizing sign language and gestures, drawing in air, playing motion based games, and even identifying plants and wildlife. If anyone is interested, I'd like to start developing an Activity that uses face/object detection for something more fun or useful. I'm new to Python and OLPC development, so I could use some help. Alternately, I am a student, so if anyone is interested in mentoring this as a possible Summer of Code project, that would be great. ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: Face Detection and other Vision Processing (and some sample code)
OpenCV has a Python API that uses SWIG. It's in the default fedoro repo as opencv-python. Good idea, I'll add it to the wiki momentarily. On Thu, Mar 20, 2008 at 6:47 PM, Chris Hager [EMAIL PROTECTED] wrote: Nirav Patel wrote: I've been playing with vision processing stuff for the XO, mostly using OpenCV. I have some sample code up at http://eclecti.cc/code/face-detection-on-the-olpc-xo that does face detection surprisingly quickly. It seems that the bottleneck isn't even the OpenCV Haar Cascade algorithm, but the time it takes to initialize the v4l2 drivers. Wow - this is amazing! And especially with this speed of ~0.25 seconds for finding a face! How are you accessing OpenCV - via SWIG or CTypes? I'd suggest to create [[OpenCV]] and start linking to infos, howto's and examples. That will make life much easier for all who want to play around with image recognition on the xo in the future. - Chris ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel