Go Down

Topic: 'Stealing' LCD output (Read 1 time) previous topic - next topic


hey guys, i was wondering, if i want to take a non contact thermometer and feed the temperature vales which it displays on the built in lcd to my pc for storage, how can that be done?


aim a webcam on it ?

You need to provide more details about the device used to give a tech answer. You might tap the signal at the sensor, or near the lcd both are possible, but it requires in both cases some serious reverse engineering.
Rob Tillaart

Nederlandse sectie - http://arduino.cc/forum/index.php/board,77.0.html -
(Please do not PM for private consultancy)


hmm that sounds incredibly complicated, the problem i see with the webcam is that picture taking may be very slow, especially since i want to use this device to make thermal area plots for a thermocam. what is the fastest image acquisuition rate i can get from a typical consumer webcam?

heres the basic concept:

-take two servos and assemble them to provide x-y control
-connect thermometer and webcom to servos, placing the webcam so that it centered at the same point the laser from the thermometer is aimed at.
-place servos at 0 displacement(looking dead centre) and take picture with cam.
-based on the FOV of the camera, change servo angle to aim where the top right corner of the first image would have been.
-*turn on thermometer for say 30ms by replacing the physical trigger with a digital one connected to the board, recording output value from lcd to the pc(along with servo angle) 2-3 times during this period.
-move servo in raster motion repeating this process to obtain a 'grid' of thermal values.
-use graphical postprocessing to shade the grid with desired colors based on temperature value at each cell in the grid
-overlay the original image with the thermal grid(for eg with photoshop) for a cheap thermo-map.

i suppose if i am using a webcam with a fast enough image capture rate i can have it take images of the lcd at the rate stated in the * step, then record the values after by examining them in a slideshow. I also guess i could program analytical software to then auto convert the values using image recognition, but this again seems like a bit of a task.

where can i learn about reverse engineering an lcd?


check - http://www.cheap-thermocam.tk/ - for inspiration

where can i learn about reverse engineering an lcd?

text only or a graphical LCD?

Course LCD hacking:
Chapter 1: find the type and typenumber, you must open the device, and some LCD's just don't have a name and number => EXIT

Chapter 2: google the datasheet, relative easy

Chapter 3: identify the communication lines, can be harder

Chapter 4: sniff the data from them, Can it be soldered? or otherwise? What protocol is used?

Chapter 5: interpret them, can also be hard but mostly ASCII characters are used, what is the data, what is overhead?.

Chapter 6: process the data, relative easy if all other succeeded

Chapter 7: answers to the exercises,  ;)

Thats about the outline, can take one hour, can take 3 weeks. 
The good part is that hacking a sensor is quite similar.

Rob Tillaart

Nederlandse sectie - http://arduino.cc/forum/index.php/board,77.0.html -
(Please do not PM for private consultancy)


hey rob, thanks for the info. i actually went to that site first, thats where i got the idea. that system will actually work just fine for my purposes, but the sensor he uses has become short supplied and relatively expensive(about 70 bucks). in comparison, a handheld therm wud run be about 40 bucks, and i would at least have the fallback of being able to just take measurements by hand if all else fails(though my data set would probably be a lot smaller due to laziness =D)

Go Up