I would like to know if you work on configrung the SSL_Vision software so that 
it works with USB-cameras.

What steps will reproduce the problem?
1. Start capture
2.
3.

What is the expected output? What do you see instead?
Run the capture in both cameras. Errors. 


What version of the product are you using? On what operating system?

SSL-Vision-r194  Ubuntu 

Please provide any additional information below.

When I start the capture ocurred this errors

CaptureDC1394vc Info: got iso channel: 9
CaptureDC1394v2 Info: using 1394B
CaptureDC1394v2 Info: running at ISO_SPEED_800!
CaptureDC1394v2 Info: Previous Packet Size: 6680
CaptureDC1394v2 Info: Recommended Packet Size: 6680
CaptureDC1394v2 Info: Final Packet Size: 6680
libdc1394 error: Error: Failed to allocate iso bandwidth
libdc1394 error: Error: Failed to setup DMA capture
CaptureDC1394v2 Error: unable to setup capture. Maybe selected combination
of Format/Resolution is not supported?

I use AVTStingray F046C and AVTStingray F033C Cameras and Pci Express Host
Firewire adapter (Two Ports)

help me please.

Regards

What steps will reproduce the problem?
1. Try to open sockets connection

What is the expected output? What do you see instead?
Should correctly open connection, but sometimes it will fall on the Else case 
and a stack smash will occur.

What version of the product are you using? On what operating system?
We are using the SSL sockets code to manage our on sockets on RoboPET, which 
runs on Ubuntu.

Please provide any additional information below.
We have found no patterns on when this problem occurs. Sometimes we are unable 
to run our software because of this problem, and we are afraid that it occurs 
on a competition.
Our team website: http://inf.ufrgs.br/pet/robopet

Dear all,

This issue was discussed before but i have not found a clear solution, I am 
currently using ubuntu 13.10 and i keep getting this error after 'make':

[ 57%] Built target sslvision
make[3]: Entering directory `/home/ssl-office/Desktop/ssl-vision/build'
make[3]: Leaving directory `/home/ssl-office/Desktop/ssl-vision/build'
make[3]: Entering directory `/home/ssl-office/Desktop/ssl-vision/build'
Linking CXX executable ../bin/client
/usr/bin/ld: cannot find -lcv
/usr/bin/ld: cannot find -lcxcore
/usr/bin/ld: cannot find -lml
collect2: error: ld returned 1 exit status
make[3]: *** [../bin/client] Error 1
make[3]: Leaving directory `/home/ssl-office/Desktop/ssl-vision/build'
make[2]: *** [CMakeFiles/client.dir/all] Error 2
make[2]: Leaving directory `/home/ssl-office/Desktop/ssl-vision/build'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/ssl-office/Desktop/ssl-vision/build'
make: *** [build] Error 2



Please provide me with clear steps of how to fix this issue! The previous 
closed issue#16 did not provide any solution!

Thanks for your concern. 

What steps will reproduce the problem?
1. Click 'neural calibration' tab on the right
2. Choose some colour. Click on some example points of that colour on the 
camera image.
3. Click on 'train neural network' and then click on 'loal neural to LUT'

What is the expected output? What do you see instead?
The ssl vision aborts.


What version of the product are you using? On what operating system?
Ubuntu 10.04

Please provide any additional information below.
Following is what is displayed at the terminal window when ssl-vision aborts.

OpenCV Error: Image step is wrong () in cvSetData, file 
/build/buildd/opencv-2.1.0/src/cxcore/cxarray.cpp, line 902
terminate called after throwing an instance of 'cv::Exception'
  what():  /build/buildd/opencv-2.1.0/src/cxcore/cxarray.cpp:902: error: (-13)  in function cvSetData

Aborted

When the two teams have robots with different heights, a workaround is needed 
in order to configure both, given the height comes from the pattern and not 
from the team. The workaround is to add a second 'standard2010' team, which is 
sort of unusual.

Using version r198

What steps will reproduce the problem?
1.I am using Ubuntu.
2. I followed the directions on: 
http://code.google.com/p/ssl-vision/wiki/RequirementsInstallation
2. when I get to this part: "After having checked out the code, the code can be 
built by executing:

make

The project should build without errors or warnings. " t
there are errors.
3. The error is "fatal error: google/protobuf/message.h: No such file or 
directory
compilation terminated."

Hello, I need help to remove detect_ball plugin, is that possible? I am using 
SSL for another application and I do not need ball detection. Thanks for your 
help

Help!

Using Eclipse to try to edit some code, I received errors in compilation time:
svn/trunk/vartypes/enhanced/TimeVar.h -> #include "TimeIndex.h"
svn/trunk/vartypes/enhanced/TimeLine.h -> #include "ClosedRange.h"

Where are these files (TimeIndex.h and ClosedRange.h, and possible others)?

Coordially,
Willytell.


Note: Eclipse is already working with CDT plugin for C/C++, Qt4 and protocol 
buffer (google's data interchange format).

What is the expected output? What do you see instead?

We have compiled and run ssl-vision with AVT Guppy cameras and started to map 
colors to the LUT. We also added a team with an included pattern. After adding 
the main areas (blue and green) of the pattern to the LUT the camera images did 
not change. That means that no thresholded images or blobs were shown.
When we change the configuration to "Read from file", the thresholdes images 
are shown and blobs are detected.
Which configuration can be made additionally so that the detection of blobs 
from camera works?

What version of the product are you using? On what operating system?

We are using the last version downloaded with svn on Ubunut Precise 12.04.2.

Please provide any additional information below.

The system was configured the following way:

Camera X -> Image Capture -> DC1394 -> Conversion Settings 

  convert to mode: rgb
  de-bayer: true
  de-bayer pattern: rggb
  de-bayer method: bilinear
  de-bayer y16 bits: 8

Camera X -> Image Capture -> DC1394 -> Capture Settings

  cam idx: X
  framerate: 60
  width: 640
  height: 480
  capture mode: raw8
  capture format: format7_0
  use 1394B: false
  use ISO800: false
  ringbuffer size: 10

Camera X -> Visualization

  enable: true
  image:  true
  thresholded: true
  blobs: true

What steps will reproduce the problem?
1. running make file.

What version of the product are you using? On what operating system?

I use Ubuntu 12.04 and the downloadable version of ssl-vision.

Please provide any additional information below.

After copying the code of Makefile,CMakeLists.txt and Doxyfile from the 
readonly version, errors about missing libraries -lcv,-lcxcore and -lml do not 
occur anymore.
But the next error that is thrown while running make is that a header file 
"ui_neurowidget.h" is not found.
Is it possible to get the newest version of ssl-vision or is there a workaround?

What steps will reproduce the problem?
1.only operate as the rules 


What is the expected output? What do you see instead?
I expected an exact position of some robot.
But I get the  projected position of the robot.

What version of the product are you using? On what operating system?
r194 ubuntu 9.10

Please provide any additional information below.

If the position which I get  is exactly the projected position,why does the 
SSL-Vision need a robot Height?
Can The SSL-Vision tell me the var "z" of the struct of SSL_DetectionBall as 
below?

ps:
message SSL_DetectionBall {
  required float  confidence = 1;
  optional uint32 area       = 2;
  required float  x          = 3;
  required float  y          = 4;
  optional float  z          = 5;
  required float  pixel_x    = 6;
  required float  pixel_y    = 7;
}

What steps will reproduce the problem?
1. Download the recorded frames from
http://www.cs.cmu.edu/~robosoccer/small/movies/ssl-vision-data/PhantomBalls.tar.
gz
2. Extract the archive to some folder
3. Load the folder from the DVR recorder
4. Use the DVR recorder to play back frames from the folder

What is the expected output? What do you see instead?
Espected output: The ball should be reported at a single static location
(from the images and the thresholded data, it seems to be quite obvious)
Output seen: In certain frames (they keep changing), "phantom" balls are
reported at locations which certainly do not have any balls or ball-like
objects. The locations of the phantom balls seem to vary between different
trials.

We are going to buy AVT Guppy F-036C Cameras and I have questions concerning 
the lenses and the camera:
Which are the fitting parameters for the lenses (focal length and f-stop)?
In which height do you mount the cameras?

Thanks in advance!

Hi, so I was following a tutorial which showed installation of SSL-vision on 
Ubuntu, but as I have Linux Mint on my laptop, I thought to try on ti first.

When I type in make at terminal, I get the following error

zaid@zp ~/ssl-vision-read-only $ make
cd build && cmake -DCMAKE_BUILD_TYPE=Release ..
Qt QTOPENGL library not found.
-- Build Type: Release
-- Arch: x86_64
-- CC Flags: -DNDEBUG -O3 -march=native -Wl,--no-as-needed
-- Custom CC Flags: 
-- Configuring done
-- Generating done
-- Build files have been written to: /home/zaid/ssl-vision-read-only/build
make -C build
make[1]: Entering directory `/home/zaid/ssl-vision-read-only/build'
make[2]: Entering directory `/home/zaid/ssl-vision-read-only/build'
make[3]: Entering directory `/home/zaid/ssl-vision-read-only/build'
make[3]: Leaving directory `/home/zaid/ssl-vision-read-only/build'
[  2%] Built target GenerateProto
make[3]: Entering directory `/home/zaid/ssl-vision-read-only/build'
make[3]: Leaving directory `/home/zaid/ssl-vision-read-only/build'
make[3]: Entering directory `/home/zaid/ssl-vision-read-only/build'
[  2%] Building CXX object CMakeFiles/sslvision.dir/src/shared/gl/glcamera.cpp.o
In file included from 
/home/zaid/ssl-vision-read-only/src/shared/gl/glcamera.cpp:22:0:
/home/zaid/ssl-vision-read-only/src/shared/gl/glcamera.h:24:19: fatal error: 
GL/gl.h: No such file or directory
 #include "GL/gl.h"
                   ^
compilation terminated.
make[3]: *** [CMakeFiles/sslvision.dir/src/shared/gl/glcamera.cpp.o] Error 1
make[3]: Leaving directory `/home/zaid/ssl-vision-read-only/build'
make[2]: *** [CMakeFiles/sslvision.dir/all] Error 2
make[2]: Leaving directory `/home/zaid/ssl-vision-read-only/build'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/zaid/ssl-vision-read-only/build'
make: *** [build] Error 2
[2]zaid@zp ~/ssl-vision-read-only $ 



The tutorial I followed is this: 
http://robosprint.com/RoboSprint2014/Ubuntu_SSL_grSim_Manual.pdf

What version of the product are you using? On what operating system?

SSL Vision 184

Windows XP


Hello everyone

I'm trying to install ssl vision in windows XP. I really appreaciate if you
have a workaround to install the software in windows computers.

Regards!!!

Hi All,

I've experienced some strange behaviors while trying to compile r198 on Ubuntu 
11.10. 

After upgrading to Ubuntu 11.10 from 11.04, I faced to many linker errors while 
trying to compile SSL-Vision. I found out that the $QT_LIBRARIES variable in 
the CMakeList.txt points to wrong place, /usr/lib instead of currect 
/usr/lib/i386-linux-gnu/. This seems like a upgrade problem/bug.

I started again with a fresh install of Ubuntu 11.10 with all SSL-Vision 
required libraries installed. This time the $QT_LIBRARIES pointed to the 
correct place, however I receive this strange error while doing `make`:

[ 60%] Building CXX object CMakeFiles/client.dir/src/client/main.cpp.o
Linking CXX executable bin/client
libsslvision.a(robocup_ssl_client.cpp.o): In function 
`RoboCupSSLClient::RoboCupSSLClient(int, std::basic_string<char, 
std::char_traits<char>, std::allocator<char> >, std::basic_string<char, 
std::char_traits<char>, std::allocator<char> >)':
robocup_ssl_client.cpp:(.text+0x27): undefined reference to 
`QMutex::QMutex(QMutex::RecursionMode)'
robocup_ssl_client.cpp:(.text+0xc3): undefined reference to `QMutex::~QMutex()'
libsslvision.a(robocup_ssl_client.cpp.o): In function 
`RoboCupSSLClient::~RoboCupSSLClient()':
robocup_ssl_client.cpp:(.text+0x127): undefined reference to `QMutex::~QMutex()'
robocup_ssl_client.cpp:(.text+0x15d): undefined reference to `QMutex::~QMutex()'
collect2: ld returned 1 exit status
make[2]: *** [bin/client] Error 1
make[1]: *** [CMakeFiles/client.dir/all] Error 2
make: *** [all] Error 2

SSL-Vision currently always sets the Firewire packet size to the "recommended" 
packet size provided by DC1394. For one of our cameras, this breaks in YUV411 
mode (a slightly smaller packet size is needed for reliable transmission). The 
attached patch allows the user to edit the Firewire packet size in the 
configuration tree, with zero meaning the recommended size.

What is the expected output? What do you see instead?
We have been modifying the (format/resolution/framerate) parameters to fit the 
camera's 
specification, but we have never been able to run the program using the DC1394 
capture setting. 
We were able however to get parameters and values selecting the "Read from 
file" capture 
setting.

What version of the product are you using? On what operating system?
We are using ubuntu 10.04 LTS, an AVT Guppy camera. According to the camera 
datasheet the 
camera supports a set of settings included in the vision program.

Please provide any additional information below.
Here's the latest entries to my Scrapbook, which includes the issues concerning 
our request:
...
• Capture Settings Parameters: Capture format :auto // use 1394B = true // 
use ISO800 true. 
OUTPUT:
    CaptureDC1394vc Info: got iso channel: 7
    CaptureDC1394     Error: unable to set operation mode. Resuming nevertheless
    CaptureDC1394 Info: using legacy 1394A
    libdc1394   Error: An ISO speed bigger than 400Mbps was requested while the
        camera is in LEGACY mode. Please set the operation mode to
        OPERATION_MODE_1394B before asking for 1394b ISO speeds
    CaptureDC1394v2 Error: unable to set ISO speed!

• Paramteros: Capture format :auto // use 1394B = false // use ISO800 false. 
Errores:
      libdc1394 error: Error: Failed to allocate iso bandwith
      libdc1394 error: Error: Failed to setup DMA capture
      CaptureDC1394v2 Error: unable to setup capture. Maybe selected combination of 
Format/Resolution is not supported?
• Testing Format/Resolution Settings:
∘ Capture format= format7_1(2,3,...,7) produces the OUTPUT: 
     Selected mode (format/resolution/framerate) is not supported by camera

Is there any sophiscated and easy way to do the colour tunning?

Yusuf

It is not possible to route data coming from ssl vision. This is a problem when 
you have a special network configuration like our team 

We applied a patch which fixes this problem / issue and is included to this 
ticket. 

Tigers Mannheim

1)graphicalClient often crashes from unknown reason (segmentation fault).
Sometimes the old ball isn't cleared from the window (the real one + the
old one that isn't exist in the packet).

2)Vartypes generates an invalid xml file from using an pointer to local object.
It causes no problem in Linux (I don't know why), the problem happens in
Windows when I tried to build an example project (vartypes).
The detail described in vartypes project.

This is not an issue when running sslvison but I think it should be fixed.
This issue may occurs in the competition :P.

What steps will reproduce the problem?
1. /ssl-vision-read-only/bin# ./vision
2.
3.

What is the expected output? What do you see instead?
QFileSystemWatcher: failed to add paths: /root/.config/ibus/bus
Bus::open: Can not get ibus-daemon's address. 
IBusInputContext::createInputContext: no connection to ibus-daemon 

What version of the product are you using? On what operating system?
ssl-vision-read-only on ubuntu 10.4

Please provide any additional information below.
I have an avt stingray connected and I am trying to run the ss-vision. The GUI 
displays but I cant get a video. Please any help will be appreciated. 

Hello everyone This is Antonio from the robobulls team I am having a bit of an 
issue. I am transferring our stuff from windows to Linux we managed to get 
everything set up vision system works and so does the graphical client.
I am trying to enhance the graphical client by adding some extra buttons but to 
even consider tackling that I first need to open and compile the graphical 
client in Qt I believe I found the correct source and after a couple of changes 
I got through a few minor details, the thing holding me back now is an error I 
do not know how to fix or get around I tried search engines but there was 
literally no results for my type of error I am hoping someone may have familiar 
knowledge of the situation or Qt the error is CentralWindow.cpp:119: error: 
'UnicodeUTF8' is not a member of 'QApplication' I have looked at the encoder 
and it is set to UTF8 that was one of the suggestions I found online aside from 
that I am stumped hope someone can assist. 

I am using QtCreator 2.62 and Qt 5.01
I have attached the file in question as well it is all files from the GClient 
provided in the SSL download package  

When trying to build using the latest ubuntu (Jaunty). I have problems
compiling. It says that Cholesky does not exist in Eigen. I have tried
setting 'EIGEN_EXTERN_INSTANTIATIONS' but that just leads to more problems.
Is there a specific version of eigen that should be used in the 2.0 line?

Creation n of an experimental branch for the 4 camera SSL-Vision and 
development towards the solution.
As requested by Tigers Mannheimm and the TC for RoboCup 2013

What steps will reproduce the problem?
1. the latest version of ssl, Ubuntu 10.04.3 LTS 
2. install all required software in README
3. make 

What is the expected output? What do you see instead?
:expected output is make without errors and warnnings,but there are some errors 
about "cannot find GL/gl.h"

What version of the product are you using? On what operating system?
: r198      ubuntu10.04.3

Please provide any additional information below.

How to use my laptop webcam as an input to ssl-vision?

The preview window does not render an image unless it is flipped 
vertically XOR horizontally. This is because backface culling is enabled. 
Backface culling is totally pointless when the only thing OpenGL is being 
used for is to draw a single 2D image into a viewport (for that matter, 
why is OpenGL even used here?)

Also, again since the only thing being rendered is a single 2D image, 
there is no point enabling alpha testing, blending, or stencil testing. 
These are just an utter waste of potential performance.

The attached patch disables everything except multisampling (multisampling 
could potentially improve graphical quality), thus fixing the bug and also 
potentially improving performance, especially for people using lower-end 
graphics hardware.