09-10-2013 07:56 AM - edited 09-10-2013 07:59 AM
Hello,
I have a LCD screen with 4 del backlight , I would like to test the backlight
Do you know the good side of this test
FYI I utlise VDM and LabWindows CVI
any idea?
This function "imaqLightMeterRect" can be used ?
Any information will be very useful for me
attached a picture to explain the principle
Thank you
09-10-2013 10:05 AM
Your illustration seems to indicate that you want to take intensity measurements in the four corners.
If your test is to check the uniformity of the entire screen, this method ill not work for you, because the intensity across the screen is non-linear.
A better way to measure the non-uniformity would be to take a series of images of the screen over a period of several minutes, starting with the screen off (cold), and then take the absolute difference between first image (coldest),last image (warmest), and all the images in between.
If you are talking about a TFT display, then the backlight intensity will improve as it warms up. Taking the abslute difference betwen images taken at different times will give you an idea about how much the intensity changes over different areas of the screen.
09-10-2013 02:27 PM
Hi,
Thank you very much for your reply
The type of my LCD is an STN LCD
See this link :
http://www.av-display.com.cn/en/Products_show.aspx?NOID=1116
We stress test time, the time should not exceed 10 sec, we have a production line and the time it is really important
Is it possible to define, for example, four areas of analysis and measure the intensity or make a color histogram for each zone and then check if there is no difference between the four areas
Can you help me
Thanks
09-18-2013 01:40 AM
09-19-2013 09:14 AM
Hi Tanite,
I think in this case your best option would be to try it out, and see how it goes! As John said, the intensity isn't going to be the same over the whole screen. So if you're checking for the entire screen to be identical, I don't think that this will work. But if the color of your lit display is significantly different from the color of the not-lit screen, and you can modify the range that it needs to be within to be considered "lit", then it is possible this implementation would work for you.
Since there's no way for us to know exactly what your screen is like without seeing it, I suggest you try setting up a quick example with a color histogram and compare the results for your 4 corners, when lit and when unlit. This will help you get a better idea of if this comparison will offer good enough results.
I hope this helps!
09-20-2013 01:30 AM