In traffic applications, an accurate global shutter is needed to handle moving objects, 60 fps is fast enough, and the stability in temperature extremes is required.ĬCD-based cameras are still relevant, but they are starting to become ideal only for niche applications. There are also many machine vision applications where a perfect global shutter is required such as measuring the liquid level in a moving container or monitoring a fast chemical reaction. In machine vision applications with constant illumination and short integration times, CCD may be a better choice. This does not mean CCD is dead, and there are still many applications that can benefit from CCD cameras. CMOS cameras are now leading in new applications such as CMOSIS-based cameras in biometrics:
The latest CMOS technology brought a much more complex pixel with correlated double sampling (CDS) for excellent image quality and a better global shutter.ĬMOS cameras have been taking over for CCD in many traditional machine vision applications such as inspection and positioning. Over the last few years CMOS image sensors have been introduced into the market with much better global shutter technology and more recently in combination with dramatic improvements in read noise and dark current related shot noise performance. CCD has a perfect global shutter, but the gap is closing. CCD still maintains an advantage in terms of dark current as CCD has very predictable dark current behavior at high temperatures. There is some focus on improvements in Qe in the NIR. For instance, the current state of the art is read noise of 9 electrons. In the last few years (fueled by investments in the consumer market), there have been dramatic improvements in sensitivity of CMOS image sensors – most importantly Qe and read noise.ĬCD technology is mature and very good, but does not have much room to get better – it is reaching the technology limits. In the 2 years since we last examined this topic there have been a lot of changes.įor many years CCD image sensors had the advantages over CMOS including a perfect global shutter, the lowest dark current, better Qe in the NIR, good MTF…basically better overall image quality.
There has been a continued improvement in CMOS and increased use in CMOS image sensors over CCD in machine vision and global security applications. However, since that is not the case at the moment (examples expected in 2015) what to do if you need to make an industrial camera selection now? Okay so we just talked about how you may not have to make a choice between a CCD or CMOS image sensor.
0 kommentar(er)
