This is a follow-up of my previous post: Setting up OpenCV 2.3 and Netbeans 7.0 on Ubuntu Oneiric 11.10. A couple of days ago Gijzelaar’s PPA for OpenCV has updated from 2.3 to version 2.3.1. So I thought I’d share this with you guys. I also think this is a slightly better approach

OpenCV (Open Source Computer Vision Library) is a library of programming functions mainly aimed at real time computer vision, developed by Intel and now supported by Willow Garage. It is free for use under the open source BSD license. The library is cross-platform. It focuses mainly on real-time image processing. (Wikipedia)

The upgrade did leave my machine with some broken packages, so be cautious! I managed to fix the broken packages, but that’s another story and it probably was a rare case.

Okay, so here is a quick guide:

1. Installation of OpenCV 2.3.1

Launch a terminal and execute these commands. This adds the PPA of our friend Gijs Molenaar, which now contains OpenCV 2.3.1.

$ sudo add-apt-repository ppa:gijzelaar/cuda
$ sudo add-apt-repository ppa:gijzelaar/opencv2.3
$ sudo apt-get update
$ sudo apt-get install libopencv-dev

2. Installation of Netbeans 7.1

Ubuntu 11.10 also comes with Netbeans 6.9, just download the Netbeans 7.1 from the Netbeans Homepage and you’re all set. Be sure you download the version with C/C++ support.

3. Set the project properties

Create a new C/C++ project in Netbeans 7.1. First, set the project properties correctly. These settings worked for me.

In the C++ Compiler dialog add “/usr/locale/opencv2″ to the “Include Directories” field.

In the Linker Dialog, Click on the “…” button for Libraries, a new Dialog appears, choose “Add PkgConfig Library…”. Search for “opencv” and add it.

This should add the following library expression

`pkg-config --libs opencv`

(Note that this is a better solution than in my previous post)

4. Set the includes

Note: this has also changed since my previous post!

#include "opencv2/highgui/highgui.hpp"
#include "opencv2/core/core.hpp" 
#include "opencv2/imgproc/imgproc.hpp"

I find it easy to work with the cv and std namespace by default, so I don’t have to type “cv::” before every function.

using namespace cv;
using namespace std;

And if we re-use our previous example (yea, I’m lazy), a little program that shows an image, you have something like this:

#include "opencv2/highgui/highgui.hpp"
#include "opencv2/core/core.hpp" 
#include "opencv2/imgproc/imgproc.hpp"
#include <iostream>
using namespace cv;
using namespace std;
 
int main(int argc, char *argv[]) {
 
    Mat image;
    image = imread(argv[1], CV_LOAD_IMAGE_COLOR);
 
    if (!image.data) // Check for invalid input
    {
        cout << "Could not open or find the image" << endl;
        return -1;
    }
 
    namedWindow("Display window", CV_WINDOW_AUTOSIZE); // Create a window for display.
    imshow("Display window", image);
 
    return 0;
}

OpenCV has really poor/decentralized documentation, which makes these simple things over-complicated. I hope this helps some people struggling with getting OpenCV to work. Let me know if it works!

Cheers!
Ruben Verhack

OpenCV (Open Source Computer Vision Library) is a library of programming functions mainly aimed at real time computer vision, developed by Intel and now supported by Willow Garage. It is free for use under the open source BSD license. The library is cross-platform. It focuses mainly on real-time image processing.

1. Installation of OpenCV 2.3

First of all, Ubuntu 11.10 comes with OpenCV 2.1, and I wanted 2.3. Luckely our friend Gijs Molenaar provides us with a PPA for OpenCV 2.3.

$ sudo add-apt-repository ppa:gijzelaar/cuda
$ sudo add-apt-repository ppa:gijzelaar/opencv2.3
$ sudo apt-get update
$ sudo apt-get install libopencv-dev

(Little notice: the package that comes with Ubuntu 11.10 is called libcv-dev, however in the PPA of Gijs the package is called libopencv-dev)

You might also want to install other packages related to OpenCV, just search for them in the Ubuntu Software Center. However, libopencv-dev will do.

2. Installation of Netbeans 7.0

Ubuntu 11.10 also comes with Netbeans 6.9, just download the new version from http://netbeans.org and you’re all set. Be sure you download the version with C/C++ support.

3. Set the project properties

Then create a new C/C++ project in Netbeans 7.0. First, set the project properties correctly. These settings worked for me.

In the C++ Compiler dialog add “/usr/locale/opencv2″ to the “Include Directories” field, also do this in the Linker dialog. In the Linker Dialog, also add the libraries opencv_core and opencv_highgui. You can find these libraries directly under “/usr/lib/”.

4. Set the includes

#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>

I find it easy to work with the cv and std namespace by default, so I don’t have to type “cv::” before every function.

using namespace cv;
using namespace std;

And if we put this all together in to a little program that shows an image, you have something like this:

#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <iostream>
using namespace cv;
using namespace std;
 
int main(int argc, char *argv[]) {
 
    Mat image;
    image = imread(argv[1], CV_LOAD_IMAGE_COLOR);
 
    if (!image.data) // Check for invalid input
    {
        cout &lt;&lt; "Could not open or find the image" &lt;&lt; endl;
        return -1;
    }
 
    namedWindow("Display window", CV_WINDOW_AUTOSIZE); // Create a window for display.
    imshow("Display window", image);
 
    return 0;
}

There we go: Our first OpenCV program in Netbeans! I hope this will work for you too

It’s been a while since I wrote something, but I liked to share this with you all. I bought a new Dell Studio 1558 and all the fn-keys were working except the one to switch the displays…

I enjoy switching between dual-screen, using only my external 22″ or using only my laptop, when I’m on the road. So I was looking for a workaround.

This laptop, is my first machine with Ubuntu that uses an ATI graphics chip, so this was all quite new to me. But I have to say; I’m very very pleased with it.

Ok, so what do you have to do to get things working?

1. Create the script
Create the script, I called it toggleDisplay.sh and this is how it looks in my case:

#!/bin/bash
# Script to toggle display configuration on Dell Studio 1558
# author: Ruben Verhack
 
config="/tmp/display.conf"
current=`cat $config`
 
# Check if config exists
if [ ! -a $config ]
then
   # Empty file executes default
   touch $config;
fi
 
# Check if CRT2 is connected
if xrandr -q | grep "CRT2 connected"
then
   # Toggle between states
   case "$current" in
   '')
      # default
      xrandr --output LVDS --auto --output CRT2 --left-of LVDS --output CRT2 --auto
      echo "dual" > $config
      ;;
   'dual')
      # was dual, now external only
      xrandr --output LVDS --off --output CRT2 --auto
      echo "external" > $config
      ;;
   'external')
      # was external, now laptop only
      xrandr --output LVDS --auto --output CRT2 --off
      echo "laptop" > $config
      ;;
   'laptop')
      # was laptop, now both
      xrandr --output LVDS --auto --output CRT2 --left-of LVDS --output CRT2 --auto
      echo "dual" > $config
      ;;
   esac
else
   xrandr --output LVDS --auto
fi

This toggles your displays from dual screen to external only to laptop only, and from there to dual screen again. Note that my monitors were called LVDS and CRT2, you can easily check how your monitors are called by executing:

xrandr -q

It’s pretty easy, huh?

2. Bind the key to the script
Make sure you given the script rights to be executed, let’s assume you’ve put the shell script in your home-folder.
chmod +x ~/toggleDisplay.sh

Next launch Gnome Keybinding by launching System > Preferences > Keybindings .

Add a custom action:
- Name: Toggle Displays
- Action: sh ~/toggleDisplay.sh

Select it, click on the shortcut and press the key you want to bind to this action and presto!

I hope this helped someone!

As a PHP developer it comes in handy to run your websites without having to upload them to your server. You can run them on your own computer if you want to, there isn’t much to it. I also like using the CodeIgniter framework for my projects, this is a really speeds up the whole process!

I will take you through the basic steps to set up your localhost to get CodeIgniter running.

1) Use Synaptic package manager to install “apache2″, “php5″ and “mysql-server”.

During the installation you will be asked to enter the password for the root user for your mysql-server.

2) Create a symbolic link, so that when you go to localhost in your browser, you get the website you want.

sudo cd /var/
sudo rm -r www/
sudo ln -s /path/to/website www

Always look before you delete something with “rm -r”, normally there should only be a index.html file in your www/ folder.

I prefer creating a link to my website that I put somewhere in my home directory, as this is being back-upped frequently, and is on another partition.



3) Install the package “phpmyadmin” through synaptic.

It is important that you wait to install phpmyadmin until you have installed the other packages and created the symbolic link. Otherwise it can be a mess to get things running.

4) To allow mod-rewrite, to get the clean CodeIgniter URL’s, you need to do the following.

sudo a2enmod rewrite
sudo nano /etc/apache2/sites-available/default

You will see something like this:

        DocumentRoot /var/www
        <Directory />
                Options FollowSymLinks
                AllowOverride None
        </Directory>
        <Directory /var/www/>
                Options Indexes FollowSymLinks MultiViews
                AllowOverride None
                Order allow,deny
                allow from all
        </Directory>

Change the AllowOverride’s to All.

5) Restart Apache

sudo /etc/init.d/apache2 restart

This wasn’t so hard, and it really makes life easier!

Note: You can’t use the function mail() or CodeIgniter’s mail-class without further configuration. I won’t go into details about this.

Also note that there are many ways to tackle this, this is only one easy suggestion. If you need to switch a lot between different websites, then changing the symbolic link each time is too much work. You can check out this article here.

Happy coding!

I did it on a fresh install of Ubuntu 9.04 RC. It has known to work on Ubuntu 8.10, but it gave me a segmentation fault, probably because I had been messing around . I tested it with the trial version downloaded from adobe.com.

To do so, you need Wine. I will take you through the steps that I’ve taken:

1. If you have used wine before, please start of with a clean .wine folder. Or bottle you Photoshop by doing every action with a specific wineprefix.
2. Install Wine 1.1.16 from here.
3. Download winetricks and install necessary libraries. Also make sure you have a decent driver for your graphics card installed.

wget http://www.kegel.com/wine/winetricks
sh winetricks msxml6 gdiplus gecko vcrun2005
sudo apt-get install msttcorefonts

3. If you have a non-english Ubuntu installed please put LANG=C before the following command like this:

LANG=C wine Setup.exe

4. Install Wine 1.1.19 from here, to fix the menu bar bug.
5. To fix the text-layer bug, put atmlib.dll in “~/.wine/drive_c/windows/system 32″. You can download it from several places if you search it on google.
6. Run it!

It was tested on an amd64 and i386.

The reason why you have to set the “LANG” variable, is that Photoshop searches for files in “Program files”, if you don’t set it, wine installs it into a local translation of Program files.

Very happy that Wine made such a big progress over the last years. But there are a lot of regressions in the last versions. That is the reason you sometimes have to try other versions of wine for the installer. But 1.1.16 and 1.1.19 are not stable versions, so I forgive them. Last version that was declared stable till today was 1.0.1.

I also got Ableton 7 and MS Office Excel, Word and Powerpoint 2007 working with wine!

Big thanks to Quitte for helping me out.

So I switched to Thunderbird. But then same symptoms started showing. Giving me 5-10 crashes of Thunderbird a day, and serious troubles on heavy browsing.

The first things to do when getting crashes of Mozilla products are:

1. Make sure you don’t have a lethal combination of add-ons
2. Check if Flash isn’t the evil spirit haunting your browser, try the Flashblock add-on
3. Make a new profile for Firefox and Thunderbird, by renaming your old configuration. You can find the configuration folders in your home. Mostly ~/.mozilla for Firefox, and ~/.mozilla-thunderbird for Thunderbird.
4. Completely remove Firefox (make a backup of your configuration) and reïnstall it

I tried every possible thing I could do to fix Thunderbird or Firefox. But then again, why was Evolution suffering as well? Well the deal was that I had WINS resolution enabled in Samba. Found it in this bug report.

When upgrading to 8.10, Samba was upgraded too. This bug has been reported and hopefully will be solved in Ubuntu 9.04, launching 23rd of April.

So how do we fix this?

Enter this in your console:

sudo nano /etc/nsswitch.conf

And look for a line like this:

hosts:          files mdns4_minimal [NOTFOUND=return] wins dns mdns4

Change it into

hosts:          files mdns4_minimal [NOTFOUND=return] dns mdns4

and everything should go back to normal.

However the WINS resolution is a good tool in Samba. I don’t really need it, because I don’t have the need to use a local network with Windows computers, but I can imagine it’s not that pleasant for people who do. Downgrading your Samba (or Ubuntu) is another option.

WINS resolution is used to communicate to Windows computers using their computer name, rather than their IP.

That’s it!

Thickbox is a very nice javascript library based on jQuery (which is also a very neat library). It is used to show images and documents in a sort of popup like Lightbox does.

Let me sketch the problem:

1) I load my page (also loading thickbox.js)
2) I change my content in one dom-element, for example a div element, through Ajax.
3) Then I want to call thickbox on an element in the new div.

You will that see nothing happens, because Thickbox hasn’t been able to register this object.

So basically we need to reinitialize Thickbox.


Notice that you should download “thickbox.js” and not “thickbox-compressed.js” for this feature.

In thickbox.js there is a function tb_init(). This is called when the script is being loaded to add the click-events to the elements which have “thickbox” as class.

Recalling this function alone is not enough! Because in some cases elements will have two or more same click-events. It is better to remove the click-events first.

Add this to your thickbox.js:

1
2
3
4

function tb_reinit(domChunk) {
	$(domChunk).unbind("click");
	tb_init(domChunk);
}

And call this function whenever you load new content through Ajax which can possibly contain new Thickbox elements.

tb_reinit('a.thickbox, area.thickbox, input.thickbox');

Last but not least: Be careful when upgrading to a new version of thickbox, This was written for Thickbox version 3.1 .

And that’s it!

Hope this was useful.

I always was sloppy with my backups and I really needed to do something about it. I bought a new Freecom HDD 500GB and what I wanted to do is quickly backup my files every time I plug the device in, and it should give me a proper notice when the backup is done.

Doesn’t sound that hard, or does it?

It became clear that I needed a little thingy called “udev”. And with big help from unutbu from ubuntuforums.org I managed to get a very nice script up and running.

First thing you have to do is to know how your system recognizes your USB drive.

I followed this great tutorial here. (Don’t bother the fstab!)

My rules look like these.

1
2

 
SUBSYSTEMS=="usb", KERNEL=="sd?1", ATTRS{product}=="Freecom Network Drive", NAME="freecomHD", RUN+="/usr/bin/usb_backup.sh"

Note that I have saved them under “/etc/udev/rules.d/81-local.rules”, because /etc/udev/rules.d/README says. It would also be safer I you select your device using UUID.

Files should be named xx-descriptive-name.rules, the xx should be
chosen first according to the following sequence points:

80 rules that run programs (but do not load modules)

After adding your rule, you need to restart udev.

$ sudo /etc/init.d/udev restart

Next install libnotify-bin through synaptic. This will make it possible for us to show notifies as if it were GNOME itself

Next we’ll need to hack the script a little bit:

Place this in “/usr/local/bin/alt-notify-send”

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

#!/bin/sh
user=`whoami`
pids=`pgrep -u $user gnome-panel`
title=$1
text=$2
timeout=$3
 
if [ -z "$title" ]; then
        echo You need to give me a title >&2
        exit 1
fi
if [ -z "$text" ]; then
        text=$title
fi
if [ -z "$timeout" ]; then
        timeout=60000
fi
 
for pid in $pids; do
        # find DBUS session bus for this session
        DBUS_SESSION_BUS_ADDRESS=`grep -z DBUS_SESSION_BUS_ADDRESS \
                /proc/$pid/environ | sed -e 's/DBUS_SESSION_BUS_ADDRESS=//'`
        # use it
        DBUS_SESSION_BUS_ADDRESS=$DBUS_SESSION_BUS_ADDRESS \
        notify-send -u low -t $timeout "$title" "$text"
done

This is because

The reason why notify-send was not working was because notify-send uses the DBUS_SESSION_BUS_ADDRESS environment variable and that variable was not set in the shell that runs your script. This environment variable is set when you start a gnome session, but is not set when root runs a udev rule.

Moreover, one user can not send messages to another user using notify-send. So we use “su” to change root to your normal user, and we use the alt-notify-send script to find and set the DBUS_SESSION_BUS_ADDRESS before launching notify-send.

(unutbu)

Then I created a file /usr/bin/usb_backup.sh.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

#!/bin/bash
{
mount_point=$(grep freecomHD /etc/mtab)
attempts=1
while [ -z "$mount_point" ] && [ "$attempts" -le 50 ]; do
    # $mount_point has not been found
    # quit if this fails more than 50 times. 
    # This should not be necessary, but it better to be safe
    # than have the script trapped in this loop forever for
    # some unforeseen reason.
    sleep 1
    mount_point=$(grep freecomHD /etc/mtab)
    attempts=$(($attempts+1))
done
 
if [ -n "$mount_point" ]; then
    su ruben alt-notify-send "Backup Message" "USB Backup device detected" 0
 
    rsync -a /home/ruben/ /media/FREECOM\ HDD/
 
    su ruben alt-notify-send "Backup Message" "Your USB Backup has completed" 0
fi
} &

Note that my user is named “ruben” and that you should replace it by your own. Also replace the the name in mount_point to the name you have given in your udev-rules.

Also note that the whole process is run in a new process (due to the “&” character). Otherwise ubuntu will try to do the backup before the mount. But it needs to be mounted before backup. This requires the little hack on top of the script

Don’t forget to make the scripts executable!

1
2

$ sudo chmod 755 /usr/bin/usb_backup.sh
$ sudo chmod 755 /usr/local/bin/alt-notify-send

It may be a good thing to tweak your rsync a little bit, so you don’t lose too much time every time you plug in.

And that’s it! This was a little harder than I thought. But then again, it shows that you’re absolutely free to do whatever you want on ubuntu. This is why we all love our tux.

Hope this helped somebody. And stick around to read more about all sorts of stuff!

Happy coding!