I have dash cams in my car. Alright so thats no big reveal. However, one thing I have noticed is that when driving at night, the rear facing camera is incapable of recording vehicle details, due to being swamped by their headlights.
My thoughts are that if the camera was supplemented by a powerful infrared illuminator, this would provide sufficient reflected IR to allow the camera to register the details of at least a vehicle close behind. The principle being much the same as using a fill flash on a still camera to bring out detail on an otherwise washed out scene.
So I obtained a cheap IR LED lamp module. This has, I discovered, a CdS photoresistor to turn it on when the light level is low.
Being IR, its not so easy to test as a visible lamp! So, I let the CCD camera in my phone take a look at its output, with the lights off (so activate the sensor)
On the bench, it certainly looks bright! I suspect that the CdS cell will need to be disabled in use, otherwise the headlights of a following vehicle with turn the IR off!
I have also today returned to the big LED message board. Now, theres no way im ever going to get the firmware for this, so my only option is to find out what data it needs to drive its panels, and recreate that using another method. Each of the panels (I think theres six), has ten LED dot matrix arrays, in two rows of five, and each LED dot matrix is an 8x8 array. Each array has identical driver arrangements consisting of several ICs and transistors. There are also other ICs which look to be responsible for the addressing of the arrays and the panels. Luckily, none are custom devices, but are all standard logic devices.
So the fun of reverse engineering one of the display 'cells' to find out the actual circuit, has begun. Already, I know several of the connections between the chips. Its just a case of lots and lots of continuity tests!