ElphelWiki - User contributions [en]

Elphel workshop in Bordeaux during RMLL 2010

2010-06-20T05:04:35Z

Luxigo: /* Deep hardware / software integration with Elphel: example on likoboard and likomapper software */

= About RMLL =

[http://2010.rmll.info/ RMLL 2010] (Libre Software Meeting) is a free (as a beer and as a speech) and non commercial conferences, workshops and round table about Free Software and its applications. The LSM goal is to provide a platform for exchange among Free Software developers, users and stakeholders.

This year RMLL will take place in Bordeaux, France from 6th to 11th of July and is organised by a set of local Free Software user groups and organizations with the help of several public and private sponsors. The event will be hosted by ENSEIRB-MATMECA, U-Bordeaux 1, LaBRI & ENSCB.

= About the workshop =

The workshop will take place during RMLL 2010 and is organized by our Swiss partner [http://alsenet.com/ Alsenet SA].

All our European customers, partners and friends of Elphel are invited to participate in the workshop. It will be a perfect occasion to meet other Elphel customers, users & developers, to test the hardware, software and to discuss project ideas and future developments. RMLL is also an ideal place to get a better understanding of the economical models of Free Software.

Some preparations on your end (e.g. pre-installing required packages) would allow us to focus on the main topics and cover any remaining questions. Any question prior or during the workshop should be discussed on Elphel’s IRC channel, wiki or public mailing list so we can have a public log that is potentially useful to other Elphel users as well.

Please sign up on the [http://doodle.com/89tms34cippzw4kv participant’s list], and please add your availability during RMLL so that we can find the optimal day & time for the workshop. All further information about this workshop will be available on this wiki page.

We are looking forward to seeing you at the workshop!

= Overview of the camera building blocks =
== boards ==
== assemblies ==

= Elphel SDK =
= reflashing camera firmware & FPGA bitstream =

= Overview of th main software available on the camera =
== IMGSRV ==
== astreamer ==
== camogm ==
== daemons ==
== lighttpd / FastCGI / PHP ==
== different PHP scripts ==

= client software compatible with Elphel cameras =
= few words about network configuration for unicast and multicast modes =

= Post-processing =
== imageJ plugins for Elphel ==
== JP46 post-processing workflow ==
== Gstreamer plugins for Elphel ==
== using Gstreamer and GLSL with Elphel cameras ==
== working with OpenCV ==
== using OpenCV and GpuCV with Elphel cameras ==

= Interfacing with the camera, triggering, synchronization =
== simple and stupid integration with Arduino ==
== trigger internal & external, multiple cameras synchronization, trigger a camera from GPS ==
== Deep hardware / software integration with Elphel: example on likoboard and likomapper software ==
=== Cross-compiling the libs ===

liblikoboard depends on libhid, and libhid needs libusb legacy.

I did compile the latest libusb legacy (0.1.12) from sourceforge,
and libhid-svn (rev. 364) from http://libhid.alioth.debian.org/

[[Cross_compiling_standalone_packages_for_the_camera|See here for the cross-compiling method]]

=== Porting the PHP extension ===

To port the PHP extension php_likoboard I had to:

* Build it for my native cpu,
* Source init_env from the elphel353/ folder.
* Copy the php_likoboard folder to elphel353/apps/php/ext/php_likoboard.
* Change to this folder and run ../../elphize.
* Run make and copy .libs/php_likoboard.so to /usr/local/crisv32/lib/php/extensions/no-debug-non-zts-20060613/.

=== Debugging ===

php_likoboard was not working on the camera. So I did write a test application sending debugging messages in console to see what was going wrong.

libusb-0.1.12 appears to be broken when running on the camera, so I did restart using libusb-0.1.11 and it worked.

=== Adding a PHP ELPHEL_DAEMON to the firmware ===

=== Integration to Elphel CVS ===

= multisensor examples with 10359 board =

Elphel workshop in Bordeaux during RMLL 2010

2010-06-20T04:56:10Z

Luxigo: /* deep hardware / software integration with Elphel: example on likoboard and likomapper software */

= About RMLL =

[http://2010.rmll.info/ RMLL 2010] (Libre Software Meeting) is a free (as a beer and as a speech) and non commercial conferences, workshops and round table about Free Software and its applications. The LSM goal is to provide a platform for exchange among Free Software developers, users and stakeholders.

This year RMLL will take place in Bordeaux, France from 6th to 11th of July and is organised by a set of local Free Software user groups and organizations with the help of several public and private sponsors. The event will be hosted by ENSEIRB-MATMECA, U-Bordeaux 1, LaBRI & ENSCB.

= About the workshop =

The workshop will take place during RMLL 2010 and is organized by our Swiss partner [http://alsenet.com/ Alsenet SA].

All our European customers, partners and friends of Elphel are invited to participate in the workshop. It will be a perfect occasion to meet other Elphel customers, users & developers, to test the hardware, software and to discuss project ideas and future developments. RMLL is also an ideal place to get a better understanding of the economical models of Free Software.

Some preparations on your end (e.g. pre-installing required packages) would allow us to focus on the main topics and cover any remaining questions. Any question prior or during the workshop should be discussed on Elphel’s IRC channel, wiki or public mailing list so we can have a public log that is potentially useful to other Elphel users as well.

Please sign up on the [http://doodle.com/89tms34cippzw4kv participant’s list], and please add your availability during RMLL so that we can find the optimal day & time for the workshop. All further information about this workshop will be available on this wiki page.

We are looking forward to seeing you at the workshop!

= Overview of the camera building blocks =
== boards ==
== assemblies ==

= Elphel SDK =
= reflashing camera firmware & FPGA bitstream =

= Overview of th main software available on the camera =
== IMGSRV ==
== astreamer ==
== camogm ==
== daemons ==
== lighttpd / FastCGI / PHP ==
== different PHP scripts ==

= client software compatible with Elphel cameras =
= few words about network configuration for unicast and multicast modes =

= Post-processing =
== imageJ plugins for Elphel ==
== JP46 post-processing workflow ==
== Gstreamer plugins for Elphel ==
== using Gstreamer and GLSL with Elphel cameras ==
== working with OpenCV ==
== using OpenCV and GpuCV with Elphel cameras ==

= Interfacing with the camera, triggering, synchronization =
== simple and stupid integration with Arduino ==
== trigger internal & external, multiple cameras synchronization, trigger a camera from GPS ==
== Deep hardware / software integration with Elphel: example on likoboard and likomapper software ==
=== Cross-compiling the libs ===

liblikoboard depends on libhid, and libhid needs libusb legacy.

I did compile the latest libusb legacy (0.1.12) from sourceforge,
and libhid-svn (rev. 364) from http://libhid.alioth.debian.org/

[[Cross_compiling_standalone_packages_for_the_camera|See here for the cross-compiling method]]

It was not working on the camera, so I dit write a test application to see what was going wrong.

libusb-0.1.12 seems broken when running on the camera, so I did restart using libusb-0.1.11 and it worked.

Then I was able to compile liblikoboard using the same method.

=== Porting the PHP extension ===

To port the PHP extension php_likoboard I had to:

* Build it for my native cpu,
* Source init_env from the elphel353/ folder.
* Copy the php_likoboard folder to elphel353/apps/php/ext/php_likoboard.
* Change to this folder and run ../../elphize.
* Run make and copy .libs/php_likoboard.so to /usr/local/crisv32/lib/php/extensions/no-debug-non-zts-20060613/.

=== Adding a PHP ELPHEL_DAEMON to the firmware ===

=== Integration to Elphel CVS ===

= multisensor examples with 10359 board =

Cross compiling standalone packages for the camera

2010-06-20T04:05:48Z

Luxigo: /* Setting environment variables */

This procedure is intended to
* test software before including it in the cvs,
* to build software that would be too big to fit into the firmware image,
* to build standalone packages you just dont want to include into the firmware image.

=First=
ELPHEL_SOURCEDIR and TOOLCHAIN values below are installation dependent, check if it match yours.

=Setting environment variables=

You must source init_env from the Elphel firmware source directory, eg:
ELPHEL_SOURCEDIR=/usr/local/src/elphel353-8.0.7/elphel353
cd $ELPHEL_SOURCEDIR
source init_env

and set the following variables:

# toolchain
TOOLCHAIN=/usr/local/cris
export PATH=/usr/local/crisv32/bin:$PATH
export PATH=${TOOLCHAIN}/crisv32-axis-linux-gnu/bin:$PATH
export PATH=${TOOLCHAIN}/bin:$PATH

# binaries
export CC=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-gcc
export CXX=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-g++
export CPP=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-cpp
export AR=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-ar
export RANLIB=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-ranlib

# options
export CFLAGS=-I${TOOLCHAIN}/include
export CFLAGS="${CFLAGS} -I${TOOLCHAIN}/crisv32-axis-linux-gnu/sys-include"
export CXXFLAGS=${CFLAGS}
export CPPFLAGS=${CFLAGS}
export LDFLAGS=-L${TOOLCHAIN}/lib
export LDFLAGS="${LDFLAGS} -L${TOOLCHAIN}/crisv32-axis-linux-gnu/lib/"

# user includes and libraries
export LDFLAGS="${LDFLAGS} -L/usr/local/crisv32/lib"
export CFLAGS="${CFLAGS} -I/usr/local/crisv32/include"
export CXXFLAGS=${CFLAGS}

(check ./configure --help and the Makefile to see if more is needed)

=When the software is packaged using autoconf and automake=
You normally just have to run:
./configure --host=crisv32-axis-linux-gnu --prefix=/usr/local/crisv32 [ other options ]
make
make install

=When the software is not packaged using autoconf and automake=
Edit the Makefile

=Install the software on the camera=
The first time, create a symbolic link for crisv32 directory in /usr/local, eg:
ssh root@cam ln -s . /usr/local/crisv32

Copy the software to the camera /usr/local directory, eg:
tar c /usr/local/crisv32/. | ssh root@cam tar xv -C /
(you may want to strip binaries and remove documentation before)

To run the software on the camera you may have to define:
export PATH=/usr/local/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

Elphel workshop in Bordeaux during RMLL 2010

2010-06-20T03:41:31Z

Luxigo: /* porting & cross-compiling the libs */

= About RMLL =

[http://2010.rmll.info/ RMLL 2010] (Libre Software Meeting) is a free (as a beer and as a speech) and non commercial conferences, workshops and round table about Free Software and its applications. The LSM goal is to provide a platform for exchange among Free Software developers, users and stakeholders.

This year RMLL will take place in Bordeaux, France from 6th to 11th of July and is organised by a set of local Free Software user groups and organizations with the help of several public and private sponsors. The event will be hosted by ENSEIRB-MATMECA, U-Bordeaux 1, LaBRI & ENSCB.

= About the workshop =

The workshop will take place during RMLL 2010 and is organized by our Swiss partner [http://alsenet.com/ Alsenet SA].

All our European customers, partners and friends of Elphel are invited to participate in the workshop. It will be a perfect occasion to meet other Elphel customers, users & developers, to test the hardware, software and to discuss project ideas and future developments. RMLL is also an ideal place to get a better understanding of the economical models of Free Software.

Some preparations on your end (e.g. pre-installing required packages) would allow us to focus on the main topics and cover any remaining questions. Any question prior or during the workshop should be discussed on Elphel’s IRC channel, wiki or public mailing list so we can have a public log that is potentially useful to other Elphel users as well.

Please sign up on the [http://doodle.com/89tms34cippzw4kv participant’s list], and please add your availability during RMLL so that we can find the optimal day & time for the workshop. All further information about this workshop will be available on this wiki page.

We are looking forward to seeing you at the workshop!

= Overview of the camera building blocks =
== boards ==
== assemblies ==

= Elphel SDK =
= reflashing camera firmware & FPGA bitstream =

= Overview of th main software available on the camera =
== IMGSRV ==
== astreamer ==
== camogm ==
== daemons ==
== lighttpd / FastCGI / PHP ==
== different PHP scripts ==

= client software compatible with Elphel cameras =
= few words about network configuration for unicast and multicast modes =

= Post-processing =
== imageJ plugins for Elphel ==
== JP46 post-processing workflow ==
== Gstreamer plugins for Elphel ==
== using Gstreamer and GLSL with Elphel cameras ==
== working with OpenCV ==
== using OpenCV and GpuCV with Elphel cameras ==

= Interfacing with the camera, triggering, synchronization =
== simple and stupid integration with Arduino ==
== trigger internal & external, multiple cameras synchronization, trigger a camera from GPS ==
== deep hardware / software integration with Elphel: example on likoboard and likomapper software ==
=== Porting & cross-compiling the libs ===

[[Cross_compiling_standalone_packages_for_the_camera|See here]]

=== Integration to Elphel CVS ===

= multisensor examples with 10359 board =

Cross compiling standalone packages for the camera

2010-06-20T03:38:47Z

Luxigo:

This procedure is intended to
* test software before including it in the cvs,
* to build software that would be too big to fit into the firmware image,
* to build standalone packages you just dont want to include into the firmware image.

=First=
ELPHEL_SOURCEDIR and TOOLCHAIN values below are installation dependent, check if it match yours.

=Setting environment variables=

You must source init_env from the Elphel firmware source directory, eg:
ELPHEL_SOURCEDIR=/usr/local/src/elphel353-8.0.7/elphel353
cd $ELPHEL_SOURCEDIR
source init_env

and set the following variables:

# toolchain
TOOLCHAIN=/usr/local/cris
export PATH=/usr/local/crisv32/bin:$PATH
export PATH=${TOOLCHAIN}/crisv32-axis-linux-gnu/bin:$PATH
export PATH=${TOOLCHAIN}/bin:$PATH

# binaries
export CC=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-gcc
export CXX=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-g++
export CPP=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-cpp

# options
export CFLAGS=-I${TOOLCHAIN}/include
export CFLAGS="${CFLAGS} -I${TOOLCHAIN}/crisv32-axis-linux-gnu/sys-include"
export CXXFLAGS=${CFLAGS}
export CPPFLAGS=${CFLAGS}
export LDFLAGS=-L${TOOLCHAIN}/lib
export LDFLAGS="${LDFLAGS} -L${TOOLCHAIN}/crisv32-axis-linux-gnu/lib/"

# user includes and libraries
export LDFLAGS="${LDFLAGS} -L/usr/local/crisv32/lib"
export CFLAGS="${CFLAGS} -I/usr/local/crisv32/include"
export CXXFLAGS=${CFLAGS}

(check ./configure --help and the Makefile to see if more is needed)

=When the software is packaged using autoconf and automake=
You normally just have to run:
./configure --host=crisv32-axis-linux-gnu --prefix=/usr/local/crisv32 [ other options ]
make
make install

=When the software is not packaged using autoconf and automake=
Edit the Makefile

=Install the software on the camera=
The first time, create a symbolic link for crisv32 directory in /usr/local, eg:
ssh root@cam ln -s . /usr/local/crisv32

Copy the software to the camera /usr/local directory, eg:
tar c /usr/local/crisv32/. | ssh root@cam tar xv -C /
(you may want to strip binaries and remove documentation before)

To run the software on the camera you may have to define:
export PATH=/usr/local/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

Cross compiling standalone packages for the camera

2010-06-20T03:30:23Z

Luxigo: /* When the software is packaged using autoconf and automake */

This procedure is intended to
* test software before including it in the cvs,
* to build software that would be too big to fit into the firmware image,
* to build standalone packages you just dont want to include into the firmware image.

=First=
ELPHEL_SOURCEDIR and TOOLCHAIN values below are installation dependent, check if it match yours.

=Setting environment variables=

You must source init_env from the Elphel firmware source directory, eg:
ELPHEL_SOURCEDIR=/usr/local/src/elphel353-8.0.7/elphel353
cd $ELPHEL_SOURCEDIR
source init_env

and set the following variables:

# toolchain
TOOLCHAIN=/usr/local/cris
export PATH=/usr/local/crisv32/bin:$PATH
export PATH=${TOOLCHAIN}/crisv32-axis-linux-gnu/bin:$PATH
export PATH=${TOOLCHAIN}/bin:$PATH

# binaries
export CC=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-gcc
export CXX=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-g++
export CPP=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-cpp

# options
export CFLAGS=-I${TOOLCHAIN}/include
export CFLAGS="${CFLAGS} -I${TOOLCHAIN}/crisv32-axis-linux-gnu/sys-include"
export CXXFLAGS=${CFLAGS}
export CPPFLAGS=${CFLAGS}
export LDFLAGS=-L${TOOLCHAIN}/lib
export LDFLAGS="${LDFLAGS} -L${TOOLCHAIN}/crisv32-axis-linux-gnu/lib/"

# user includes and libraries
export LDFLAGS="${LDFLAGS} -L/usr/local/crisv32/lib"
export CFLAGS="${CFLAGS} -I/usr/local/crisv32/include"
export CXXFLAGS=${CFLAGS}

(check ./configure --help and the Makefile to see if more is needed)

=When the software is packaged using autoconf and automake=
You normally just have to run:
./configure --host=crisv32-axis-linux-gnu --prefix=/usr/local/crisv32 [ other options ]
make
make install

=When the software is not packaged using autoconf and automake=
Edit the Makefile

Cross compiling standalone packages for the camera

2010-06-20T03:23:51Z

Luxigo: /* Setting environment variables */

This procedure is intended to
* test software before including it in the cvs,
* to build software that would be too big to fit into the firmware image,
* to build standalone packages you just dont want to include into the firmware image.

=First=
ELPHEL_SOURCEDIR and TOOLCHAIN values below are installation dependent, check if it match yours.

=Setting environment variables=

You must source init_env from the Elphel firmware source directory, eg:
ELPHEL_SOURCEDIR=/usr/local/src/elphel353-8.0.7/elphel353
cd $ELPHEL_SOURCEDIR
source init_env

and set the following variables:

# toolchain
TOOLCHAIN=/usr/local/cris
export PATH=/usr/local/crisv32/bin:$PATH
export PATH=${TOOLCHAIN}/crisv32-axis-linux-gnu/bin:$PATH
export PATH=${TOOLCHAIN}/bin:$PATH

# binaries
export CC=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-gcc
export CXX=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-g++
export CPP=${TOOLCHAIN}/bin/crisv32-axis-linux-gnu-cpp

# options
export CFLAGS=-I${TOOLCHAIN}/include
export CFLAGS="${CFLAGS} -I${TOOLCHAIN}/crisv32-axis-linux-gnu/sys-include"
export CXXFLAGS=${CFLAGS}
export CPPFLAGS=${CFLAGS}
export LDFLAGS=-L${TOOLCHAIN}/lib
export LDFLAGS="${LDFLAGS} -L${TOOLCHAIN}/crisv32-axis-linux-gnu/lib/"

# user includes and libraries
export LDFLAGS="${LDFLAGS} -L/usr/local/crisv32/lib"
export CFLAGS="${CFLAGS} -I/usr/local/crisv32/include"
export CXXFLAGS=${CFLAGS}

(check ./configure --help and the Makefile to see if more is needed)

=When the software is packaged using autoconf and automake=
You normally just have to run:
./configure --host=crisv32-axis-linux-gnu --prefix=/usr/local/crisv32 [ other options ]
make
make install

and mount or copy the resulting /usr/local/crisv32 to the camera
(you may want to strip binaries and remove documentation before)

Then to run the software on the camera you may have to define:
export PATH=/usr/local/crisv32/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/crisv32/lib:$LD_LIBRARY_PATH

=When the software is not packaged using autoconf and automake=
Edit the Makefile

File:Kifu numbering.jpg

2009-06-30T11:00:00Z

Luxigo:

File:Kifu rectify2.jpg

2009-06-30T10:59:22Z

Luxigo:

2008-12-14T22:34:50Z

Luxigo:

2008-09-15T14:38:41Z

Luxigo:

AVLD is a video loopback device for GNU-Linux, written by Pierre Parent and licensed under GPL.

With AVLD you can use Elphel cameras for image acquisition in v4l (version 1) compatible applications, video conferencing, etc

Video is playing smoothly on a 2.4Ghz dual core notebook at 25fps in 1440x896,

and CPU usage is 'only' 20-25% per core

I could not test beyond the screen and the specific camera/sensor capabilities :)

== Accessing Elphel camera streams through a v4l device (eg: /dev/video0) ==

After loading the AVLD module:

modprobe avld width=1440 height=896 fps=0

You just have to feed the video device:

mencoder rtsp://192.168.0.9 -nosound -ovc raw -vf format=bgr24 -of rawvideo -o /dev/video0

(or /dev/video1 if you had already another video device driver loaded before)

And you can play the stream for testing:

mplayer tv:// -tv driver=v4l:device=/dev/video0:outfmt=rgb24 -cache 8192 -vo xv

You can change the device parameters when module is already loaded with:

echo "width=800 height=600 fps=0" > /dev/video0

== Helper scripts ==

Note: To run those scripts you must patch avld/video_driver.c and rebuild the module before, using the patches found on the section below.

http://community.elphel.com/files/avld/avldfeed

http://community.elphel.com/files/avld/avldplay

For skype, use avlfeed options WIDTH=640 HEIGHT=480 FPS=25 FORMAT=uyvy FILTER="scale=640:480",
but generally FORMAT is bgr24

Usage: avldfeed [ -h ] [ <configuration_file> ] [ <OPTION=value> ... ]

OPTIONS:

URL=<value>
DEVICE=<value>
WIDTH=<value>
HEIGHT=<value>
FPS=<-1|0|value>
FORMAT=<bgr24|uyvy>
PALETTE=<0|1> (overriden by FORMAT)
SCALE=<none|qntsc|qpal|ntsc|pal|sntsc|spal>
FILTER=<none|mencoder_filter_chain>

avldfeed send the specified video stream to the avld video device.

Previous parameters are saved in /home/guest/.avldfeed so that you can
run the command again with no parameters or a subset, or with a
configuration file as first command argument (it will
replace /home/guest/.avldfeed)

== avld-0.13 patches ==

http://community.elphel.com/files/avld/avld_0.13-VIDIOCGWIN_null_framesize+skype_compatibility+palette_change.patch

(apply with "patch -P0 < filename.patch" in the avld sourcecode directory)

Opencv v4l capture initialization method was receiving a null frame size when it was run before watching once the stream with mplayer which is using an alternative initialization method.

I found a patch for skype compatibility on http://doc.ubuntu-fr.org/webcamvirtuelle

And I did add "palette" to the list of the avld module expected arguments.

'''It breaks compatibility with older versions''' since "palette=0|1" MUST be specified with the other parameters when loading the module or when changing the parameters writing to the video device. But it avoid scanning twice the video buffer and remove the necessity to unload the module when the palette must be changed.

AVLD homepage: http://allonlinux.free.fr/Projets/AVLD/

2008-09-06T02:37:11Z

Luxigo:

Kifu: Go game record (kifu) generator

2008-09-06T02:36:25Z

Luxigo:

'''Goban and Grid Detection:'''

1. The input image is filtered and thresholded and a big quadrilateral shape is selected, sorting contours detected with cvFindContours() from OpenCV.

[[Image:kifu_region.jpg]]

2. The selection is rectified with cvGetPerspectiveTransform() and cvWarpPerspective().

[[Image:kifu_rectify.jpg]]

3. cvHoughLines2() find lines in the rectified selection and we compute the intersections for each detected line segment.

[[Image:kifu_intersections.jpg]]

4. The intersection is validated when the angle formed by the intersection point and the line extremities is near 90 degrees; and other detected intersections that are less distant than ~92% of the expected grid spacing are discarded.

[[Image:kifu_intersections2.jpg]]

5. Intersections can be mapped back to the pre-warped coordinates and vote for the original point, and we restart at step 2 with the quadrilateral found with the next threshold value.

[[Image:kifu_backproj.jpg]]

6. The grid map is built using original point vote results, missing intersections must be interpolated.

7. Finally, we can validate the grid, or restart at step 1 with another threshold, or rectify the image again with the exact grid corners coordinates known.

'''Image change detection:'''

Using vertical and horizontal RGB components sums of a thresholded difference between a reference image and the current one,

it is easy to compute the image coordinates of a played stone

[[Image:kifu_threshold.jpg]] [[Image:kifu_sums.jpg]]

'''Stone detection:'''

When a stone is played it overlaps 1 grid square on a corner, 2 on the borders and 4 in the center.

Computing horizontal and vertical pixel sums for each grid cell or around each intersection
can tell where the stone is played and reveal the color of the stone,
being darker or brighter than the empty intersection region.

The intersection can also be seen on the difference image when the stone is white,
that could also be used to detect the stone color.

----

'''Methods for mapping the coordinates:'''

"2.2 Perspective transformation with two vanishing points"
(pages 2 and 3, equations 7 and 10)

http://cipa.icomos.org/fileadmin/papers/potsdam/2001-21-gf01a.pdf

Inverse homography and plane image rectification (page 14)

http://www-prima.imag.fr/jlc/Courses/2002/DEA-IVR.VO/DEA-IVR.VO.S2.pdf

"Inferring Projective Mappings" (page 3)

http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.9.7803

with related java and C source code here: http://www.developpez.net/forums/showthread.php?t=591698

----

'''Links:'''

irc://irc.freenode.net/#kifu

http://www.sourceforge.net/projects/kifu

http://opencvlibrary.sourceforge.net/

Photointerpretation and Small Scale Stereoplotting with Digitally Rectified Photographs with Geometrical constraints:
http://cipa.icomos.org/fileadmin/papers/potsdam/2001-21-gf01a.pdf

Vision par Ordinateur:
http://www-prima.imag.fr/jlc/Courses/2002/DEA-IVR.VO/DEA-IVR.VO.S2.pdf

Projective Mappings for Image Warping:
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.9.7803

http://sciences.ch/htmlfr/geometrie/geometrieprojective01.php

Kifu: Go game record (kifu) generator

2008-09-03T06:38:40Z

Luxigo:

'''Goban and Grid Detection:'''

1. The input image is filtered and thresholded and a big quadrilateral shape is selected, sorting contours detected with cvFindContours() from OpenCV.

[[Image:kifu_region.jpg]]

2. The selection is rectified with cvGetPerspectiveTransform() and cvWarpPerspective().

[[Image:kifu_rectify.jpg]]

3. cvHoughLines2() find lines in the rectified selection and we compute the intersections for each detected line segment.

[[Image:kifu_intersections.jpg]]

4. The intersection is validated when the angle formed by the intersection point and the line extremities is near 90 degrees; and other detected intersections that are less distant than ~92% of the expected grid spacing are discarded.

[[Image:kifu_intersections2.jpg]]

5. Intersections can be mapped back to the pre-warped coordinates and vote for the original point, and we restart at step 2 with the quadrilateral found with the next threshold value.

6. The grid map is built using original point vote results, missing intersections must be interpolated.

7. Finally, we can validate the grid, or restart at step 1 with another threshold, or rectify the image again with the exact grid corners coordinates known.

'''Image change detection:'''

Using vertical and horizontal RGB components sums of a thresholded difference between a reference image and the current one,

it is easy to compute the image coordinates of a played stone

[[Image:kifu_threshold.jpg]] [[Image:kifu_sums.jpg]]

'''Stone detection:'''

When a stone is played it overlaps 1 grid square on a corner, 2 on the borders and 4 in the center.

Computing horizontal and vertical pixel sums for each grid cell or around each intersection
can tell where the stone is played and reveal the color of the stone,
being darker or brighter than the empty intersection region.

The intersection can also be seen on the difference image when the stone is white,
that could also be used to detect the stone color.

----

'''Methods for mapping the coordinates:'''

"2.2 Perspective transformation with two vanishing points"
(pages 2 and 3, equations 7 and 10)

http://cipa.icomos.org/fileadmin/papers/potsdam/2001-21-gf01a.pdf

Inverse homography and plane image rectification (page 14)

http://www-prima.imag.fr/jlc/Courses/2002/DEA-IVR.VO/DEA-IVR.VO.S2.pdf

"Inferring Projective Mappings" (page 3)

http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.9.7803

with related java and C source code here: http://www.developpez.net/forums/showthread.php?t=591698

----

'''Links:'''

irc://irc.freenode.net/#kifu

http://www.sourceforge.net/projects/kifu

http://opencvlibrary.sourceforge.net/

Photointerpretation and Small Scale Stereoplotting with Digitally Rectified Photographs with Geometrical constraints:
http://cipa.icomos.org/fileadmin/papers/potsdam/2001-21-gf01a.pdf

Vision par Ordinateur:
http://www-prima.imag.fr/jlc/Courses/2002/DEA-IVR.VO/DEA-IVR.VO.S2.pdf

Projective Mappings for Image Warping:
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.9.7803

http://sciences.ch/htmlfr/geometrie/geometrieprojective01.php

AVLD - Another Video Loopback Device

2008-08-30T08:18:49Z

Luxigo:

AVLD - Another Video Loopback Device

2008-08-30T02:39:12Z

Luxigo:

With AVLD you can use Elphel cameras for image acquisition in v4l compatible applications, video conferencing, etc

== Accessing Elphel camera streams through a v4l device (eg: /dev/video0) ==

After loading the AVLD module:

modprobe avld width=640 height=480 fps=25

You just have to feed the video device with:

mencoder rtsp://192.168.0.9 -nosound -ovc raw -vf format=bgr24 -of rawvideo -ofps 25 -o /dev/video0

(or /dev/video1 if you had already another video device driver loaded before)

AVLD Homepage: http://allonlinux.free.fr/Projets/AVLD/

File:Kifu intersections2.jpg

2008-08-27T16:41:03Z

Luxigo: