How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!
Normally, when pixels are being read out they get moved one row down and, sequentially, one pixel at a time gets moved over and read. “With the EMCCD, instead of just being sequentially moved over and immediately read, they’re actually sequentially moved into an extended register where very high voltages are applied,” Sharma explains. “So instead of a small voltage being applied just to move the electrons over, you get a very high voltage swing applied and the electron undergoes a process of what’s called impact ionization, where it’s moved over pretty violently into the next pixel, and because of that, there is a possibility that the signal electron will generate what we call a secondary electron from it. So then you’re getting that multiplying process. You get one, it turns into two…
“There are over 500 extra pixels where this multiplication can happen. There’s only a small percentage chance that it’ll happen, but because there are so many of these extra pixels that it quickly multiplies up and you end up with that single electron—or whatever electrons were in the pixel—being multiplied into a lot more before they’re actually read out. So the way it actually works is a high voltage extended register in the CCD device that causes a multiplication of the electrons.”
A second problem, he adds, is that because such a huge amount of multiplying is necessary the dynamic range is not as high as it can be. That means a researcher can’t see the difference between dim objects and brighter objects because the image looks very saturated and bright. EMCCD, on the other hand, has a much higher dynamic range and 95 percent quantum efficiency in the visible wavelength.
The researcher also can ask the camera—Photometrics Evolve—for how many times gain he or she would like. That means anywhere from 10X to 100X up to 1,000X.
At present, the technology only works with CCD sensors. CMOS doesn’t have the type of low read noise or high quantum efficiency of EMCCDs. “As yet, we haven’t seen any technology with clear data to show that it will be better and detecting and quantifying than these cameras, but we’re always looking out for it. CMOS offers faster speed in general, but not the type of detection capability that the high-end scientists need,” Sharma says.