Using tablet PCs for processing data and displaying results

[Atul Asthana]

Present day tablet PCs (7" and even 10") are cheap, easily available and extremely versatile. (http://www.diytrade.com/china/4/prod...1/Android.html)

It will be a good idea to use these devices for processing, target identification / classification, user interaction and information display.

Thus, the microcontroller part can be used for controlling tx, rx and should be able to communicate with tablet PC.

Tablet PC can be programmed to process the data and display.

If the microcontroller part has more than one form of tx, rx methodology (both PI and VLF or multiple types of PI or mutiple types of VLF); the tablet PC could be used to automatically switch to most desired (or multiple) methodologies.


Disclaimer : I am not associated with DIYtrade or any manufacturer or distributor or supplier of any tablet PCs.

[Tec]

Yes, 3d plotting

Yes, tablets (or android phones) are a good interface..

I have been plotting 3d plots using free software (that will run on a tablet pc)... I posted it on another topic, but my link doesn't work, so I will try again...

I used 'ImageJ' to make the graphs.


Regarding the interface between the tablets and the electronics...I have been using USB,..but bluetooth modules will be more elegant. and just as easy.

The software to develop on Android (Eclipse IDE with ADK (android development Kit) is free...everything is free .

[Atul Asthana]

Great work.

[Auftragssucher]

YES!

http://www.metageo.de/html/metalog-c.html

[Alexismex]

Hello Tec, very interesting your experiment with android, can you give us more info on the electronic interface (what module you use? )between the input usb of the tablet and your PI because i want to begin some test also.
Thanks
Alexis.

[mikebg]

Hello Tec, I can not understand in what part of the plot appears the difference between ferrous and nonferrous target. Is your bottle cap ferrous?
Please make experiments with separate TX and RX coils placed in induction balance.

[Tec]

Hi Mike,

No... there is no obvious difference between the Fe and non Fe on these charts THe bottle Cap is just like a small coin...it has a quicker response. The Copper coin gets stronger with longer sample delay and longer Pulse as you probably know.

I expect there may be some difference in Fe at much longer sample delay maybe.

Yes I should do induction balance.
But here is the thing.. why do people use mono coils ?, i.e the latest minelab detectors sometimes can use mono coils... there must be an advantage.
And if Fe discrimination is possible,..why are there '5 levels of Fe discrimination' on top-of- the-range detectors, and they still can't manage to achieve reliable discrimination.

I am under the impression that mono coil gives the best depth....
I don't want to loose depth...

[Tec]

Hi Alexismex,

I am using a serial port (UART) to talk to the processor. There is a serial to USB converter (FT232RL) so I can use a USB port.

If you want to use Bluetooth, you could use a UART to bluetooth interface.

[Atul Asthana]

artificial neural networks for target clasification

It should be possible to run neural networks on a tablet to learn and classify targets, after processing raw data from the MD's TxRx.

Fast Artificial Neural Network Library (FANN) implements multilayer artificial neural networks in C. It is cross-platform, easy to use, versatile, well documented and fast. C++, PHP, PERL, Python, Delphi, .NET, Mathematica bindings and a GUI is available.

Tx -> Rx -> signal processing (tablet) -> neural network (tablet) -> human being

It should also be possible to use statistical analysis packages to carry out rote analysis of Rx data on a tablet.

The GNU Scientific Library (GSL) is a numerical library for C and C++ programmers. It is free software under the GNU General Public License.

The library provides a wide range of mathematical routines such as Eigensystems, statistics, special functions and least-squares fitting. There are over 1000 functions in total with an extensive test suite.

[mrdetector]

android software

where di i get the download ?

[Atul Asthana]

Download links for GSL and FANN

Great to see the interest. Please do let all of us know the efforts and results.

Downloading GSL

The current version is GSL-1.15. It was released on 6 May 2011. Details of recent changes can be found in the NEWS file. This is a stable release.
GSL can be found in the gsl subdirectory on your nearest GNU mirror http://ftpmirror.gnu.org/gsl/.


Fast Artificial Neural Network Library
http://sourceforge.net/projects/fann...atest/download

[Atul Asthana]

classification of metal detector signals using the wavelet transform

Feature extraction and classification of metaldetector signals using the wavelet transform and the fuzzy ARMTAP neural network

(landmines detection)

http://wenku.baidu.com/view/4ec776ea...2bcd10de0.html

[Atul Asthana]

Adaptive Resonance Theory Based Neural Network Approach for Signal Discrimination in VLF Metal Detec

Adaptive Resonance Theory Based Neural Network Approach for
Signal Discrimination in VLF Metal Detectors

ABSTRACT
This paper proposes a novel technique to discriminate among different alloys using a very low frequency (VLF) type metal detector used in typical demining operations. The typical commercially available metal detectors do not possess the ability to distinguish between normal metal debris found in a battlefield and the landmines. Statistics show that landmines cause only one out of hundred to thousand alarms in a manual demining operation, depending on the soil type and debris. Therefore, the probability of getting a false alarm is around 99% - 99.9%. The proposed method will improve the efficiency of humanitarian demining by minimizing this false alarm rate.

http://www.elect.mrt.ac.lk/V_Ara_iee.pdf

[Aziz]

Oh Boys!,

this is the end of <you know who>.

Aziz

PS: I'm waiting to the convenient Tablet PC's (smaller ones, <= 7 inch).
Best candidates:
Android or Windows O/S.
(must support USB ports)

Don't ask for iAnything (iPad, iPhone, iTab, .. ) .

[WM6]

Therefore, the probability of getting a false alarm is around 99% - 99.9%. The proposed method will improve the efficiency of humanitarian demining by minimizing this false alarm rate.

Great approach Atul Asthana. Wish you full success.