The TEPCO cam 4 at Fukushima has been having odd black-out problems. What is odd is the way that some pixels remain active while others go dark, as illustrated by these 3 screenshots taken February 12 2015, around 7:40 AM (TEPCO time stamp)
Can you see there are some active pixels (I don't know the jargon sorry) in the dark screenshot above? In addition to the blue spot in the right corner, there are hints of pixels of color in the location of the large white crane.
The image resolved and then it blacked out again partially, as illustrated in this third screenshot:
Either the cam is breaking down from cummulative radiation exposure, or there are episodes of atmospheric contamination so intense that electronics, even located at a distance, are affected.
I don't pretend to know the answer. Perhaps both factors are operative. I am simply documenting my observations and reflections.