Wednesday, July 11, 2018

A Signature Image for Documents

While filling forms and preparing documents on the computer, I come across needs to put my handwritten signature on the document. In such cases, the printing of the document, signing it by hand and then scanning it to produce the soft copy is too much. It is useful to have an image of the signature which can be easily inserted into documents.

On Linux, we can use the following steps to produce an image file where the handwritten signature stays in a transparent background.

(1) Put the signature on a white paper and take a photograph.

(2) Use a suitable tool to crop the photograph file to remove the unnecessary edges and have only the area covering the signature (e.g., Shotwell Viewer). Let's say the resulting image file is original.jpg.

(3) Run the following command to push the whiter pixels to fully white and darker pixels to fully black.

convert original.jpg -colorspace gray -threshold 28% filtered.png

(4) Run the following command to set the while pixels as transparent.

convert filtered.png -transparent white signature.png

Now, the resulting signature.png file contains the handwritten signature in black color while the background stays as transparent. This file can be used for the purpose of inserting the signature into electronic documents.

Sunday, June 17, 2018

Low Graphics Mode Problem on Ubuntu 16.04 LTS

My laptop, which runs Ubuntu 16.04 LTS, recently started to give me a strange error. When the computer is turned on, it boots into a screen where a message is shown as "Your system is running on low graphics mode". If I proceed to boot the system, the system boots up however the on-board WiFi is not detected. My computer suffered a physical damage lately which may have caused some hardware to malfunction. Anyway, being unable to fix this strange error, I found a somewhat workaround for it in the web.

Following the steps shown below to workaround each time when the "Low Graphics Mode" error occurs on Ubuntu 16.04.

(1) Turn on the computer and wait till it lands on the "Your system is running on low graphics mode" error.

(2) Press the key combination Ctrl + Alt + F1 and login to the virtual terminal it provides.

(3) On this terminal, type the following command.

sudo service lightdm restart

(4) Now, when the GUI login prompt appears, login there and the desktop environment should load now without any issues.

Cheers!

Saturday, April 28, 2018

Installing SETI@home on Ubuntu 16.04 LTS

Searching for alien life is the goal of many space exploration projects for many years. Aliens can come in all shapes and sizes starting with single celled life. Among them, finding intelligent life would be exciting than anything else. Search for Extraterrestrial Intelligence (SETI) is a project which attempts to achieve that goal by listening to their radio transmissions. They have large radio telescopes which collect radio signals from the outer space and then process them to look for any sign of message from an intelligent beings.

Now, here's the issue. The scientists at SETI don't have enough processing power in their computers to process all these data acquired from the radio telescopes. This is obviously a computationally intensive task. That's where we can give a little hand as a good citizen of Earth. We can share a part of our personal computers processing power to process SETI data. I'm writing down the steps to install the required software on a computer running Ubuntu 16.04 LTS and setup everything.

(1) First of all, we need to have an account in the SETI@home website, in order to contribute. Go to the following website and create an account.
 

(2) Now, it's time to install the relevant software on our computer. For a computer running Ubuntu 16.04 LTS, we can easily install the software through the software repositories. Open a terminal and issue the following command for that.

sudo apt update
 
sudo apt install boinc

(3) Open the BOINC software GUI and log into your account. Once you are logged in, you should be able to search for the SETI@home project from the list of projects available and add it.

(4) After a while, the SETI@home client will start taking data from the sever and process them locally. You can view the ongoing tasks on your computer in the GUI client.



Cheers!

Monday, March 19, 2018

SSIM to Measure Image Similarity

There are situations where we need to measure the similarity between two images. For example, when we have an original image and few other images with lower quality, we might need to identify when one is mostly similar to the original image. Bit-wise comparison of the image pixels is not applicable in this kind of scenarios. We need something more sophisticated. Structural similarity index (SSIM) is such a method. It provides a value between 1 and 0 when comparing two images. If two images are exactly similar the SSIM becomes 1. If they are totally different, SSIM becomes 0.

Following Python program implements SSIM to compare between to images. Since we need some extra python libraries for this code to run, we should install following libraries on Ubuntu Linux.

sudo apt install python-skimage

sudo apt install python-opencv

SSIM calculation code is as follows.


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
from skimage.measure import structural_similarity as ssim
import cv2

print("Reading files...")
first = cv2.imread("image1.png")
second = cv2.imread("image2.png")

print("Resizing files...")
first = cv2.resize(first, (2576,1125))
second = cv2.resize(second, (2576,1125))

print("Converting files to grayscale...")
first = cv2.cvtColor(first, cv2.COLOR_BGR2GRAY)
second = cv2.cvtColor(second, cv2.COLOR_BGR2GRAY)

print("Calculating SSIM value...")
s = ssim(first, second)

print("first vs second", s)

~*******************~


Thursday, March 1, 2018

Merging Video Files on Linux Terminal

When we need to merge multiple video files into a single file, there's an interesting tool called mkvtoolnix we can us. We can install it using the Linux software repository as follows.

sudo apt install mkvtoolnix

Let's say we have three mp4 video files which we need to merge and build a single mkv file. We can do it in the following way.

mkvmerge -o full-video.mkv part-1.mp4 \+ part-2.mp \+ part-3.mp

Meanwhile, I learned that there's a GUI version of the tool called mkvtoolnix-gui which I should explore later. For the moment, we can install it using the following command.

sudo apt install mkvtoolnix-gui

Cheers!

Kazam Screen Recorder

Recently, when I wanted to create a video demonstration which involves recording my computer screen and the input from my microphone. As an Ubuntu GNU/Linux user, I've tried few different screen recorder software but, the results were not very encouraging. However, in my recent attempt, I found a nice piece of software which could meet my requirements. It's called Kazam screencaster program. I just thought to leave a note about it here so that I can find it when I need it the next time.

It is available on the Ubuntu software repositories and we can install it with the following command.

sudo apt update
sudo apt install kazam




Tuesday, February 6, 2018

Running TempestSDR on Windows 10

TempestSDR is a nice tool which can be used to eavesdrop on computer monitors using the electromagnetic (EM) emissions causes by them. While this concept is not a new thing, the use of cheap software defined radio (SDR) hardware has enabled the possibility of performing this attack a lot more easier. I was struggling to get the tool up and running on Ubuntu Linux for a while and ended up without a result. Finally, I moved into Windows platform and tried it. Luckily, things went so smoothly and I got TempestSDR tool running with both RTL-SDR and HackRF hardware.

In this post, I'm writing down the steps I followed to get TempestSDR running on Windows 10 operating system with both RTL-SDR and HackRF hardware.

Preparing RTL-SDR and HackRF hardware:

It is necessary to have the required drivers installed on Wondows 10 in order to use both RTL-SDR and HackRF devices. Therefore before everything, let's get the drivers installed. A previous post written by me describes the required steps for. Refer it and install the drivers described here: http://recolog.blogspot.ie/2018/02/installing-drivers-for-rtl-sdr-and.html

Identifying the EM emission frequency of a target monitor:

Before we prepare the TempestSDR tool to eavesdrop on a computer monitor, we need to identify the frequencies where EM emissions occur on the target. This is a little bit cumbersome task as we have to go through the spectrum and identify them. We'll use SDR# software for this purpose.

(1) Download SDR# from here.
https://airspy.com/download/

(2) Extract the ZIP archive, and then inside it, double-click on the install-rtlsdr.bat file. A CMD prompt will start and download some files. It will exit automatically.

(3) Now double click on the SDRSharp.exe tool and it will open the window. You can select the "RTL-SDR (USB)" option for the source.

(4) Keep scrolling while looking for a signal which varies the peaks when I make any change in the screen of the computer such as maximizing / minimizing windows, etc. If there's a strong signal which changes the amplitude when a window is maximized, there's a good chance that it is an emission from the monitor. Note down such frequencies.

Setting up TempestSDR software:

(1) Installed JDK 8 - 32-bit version. I downloaded it from here,
http://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html The exact file I downloaded is jdk-8u151-windows-i586.exe

(2) Download and install MinGW and MSYS. We have to download a single installer and inside it, we can select the packages of MingGW and MSYS that we want to install.
https://sourceforge.net/projects/mingw/files/

(3) Set the bin folder paths to MinGW and MSYS in Windows PATH. The instructions to set PATH environmental variable on Windows can be found here: https://www.computerhope.com/issues/ch000549.htm

In my system, the paths to the bin folders of those tools after the installation were as follows.

C:\MinGW\bin
C:\MinGW\msys\1.0\bin


(4) Add JAVA_HOME path variable too. The instructions to do this can be found in this link: https://www.mkyong.com/java/how-to-set-java_home-on-windows-10/
 In my system, the path to the directory where Java was installed is as follows.

C:\Program Files (x86)\Java\jdk1.8.0_151

(5) Download TempestSDR from here. Then extract the files.

https://github.com/rtlsdrblog/TempestSDR

(6) In the very first makefile, remove the following line

@$(MAKE) -C TSDRPlugin_Mirics/ all MIRICS_HOME=$(MIRICS_HOME)

(7) Due to the fact that there are spaces in the path to Java installation directory, TempestSDR tool faces some difficulties while running the make file. Therefore, let's copy java installation folder to a new place which does not have spaces in the path.

I copied "C:\Program Files (x86)\Java" folder to "C:\Java" location.

(8) Now go into TempestSDR folder from CMD prompt and and run the following command.

make all JAVA_HOME=C:\Java\jdk1.8.0_151

If the compilation completes successfully, we are good to go.

Running the TempestSDR software:

(1) Connect either RTL-SDR dongle or HackRF device into a USB port of the computer.

(2) Go to the JavaGUI folder in the TempestSDR source code directory. There should be a jar file which we need to run.

java -jar JTempestSDR.jar

(2) From the File menu, select the "Load ExtIO source" option. Then browse to the installation directory of HDSDR software where you copied the ExtIO DLL driver for either RTL-SDR or HackRF. Select that DLL file.

(3) Select the resolution and refresh rate of the monitor being eavesdropped. Then, select the frequency of EM emanation which we manually found using SDR# software. Click "Start" and we are good to go.

Trouble Shooting:

Time to time, TempestSDR tool faced difficulties in detecting the RTL-SDR or HackRF device. In such situations, I used the following steps to resolve the issue.

(1) Restart the machine.

(2) Run SDR# with RTL-SDR/HackRF first to get the correct driver running.

(3) Then try running TempestSDR jar file from the beginning.

Following are some of the screenshots of my attempts.

A checker board image was placed on the target computer screen.

TempestSDR capturing data from a Dell monitor with RTL-SDR
 
TempestSDR capturing data from a Samsung monitor with HackRF


~*******~