Mostly, CCD camera is used for speckle measurements. I am interested to do measurements with cmos camera, are there any characteristics of cmos camera which can effect my measurements?
e.g a lower sensitivity due to the smaller fill factor, higher temporal noise, higher pattern noise, higher dark current, and the nonlinear characteristic curve are primarily mentioned.
Don't forget the advantages: When compared to CCD, a big advantage of CMOS cameras is that they are much less prone to blooming (overflow crosstalk between a highly saturated pixel and its neighbors). This is of high value when working with speckles. They also can be readout non-destructively several times, which may help you increase dynamic range. Furthermore you can easily gain framerate by not reading-out pixels outside your ROI.
I thank this is very important work .
I hope for you a nice day
Qusay Kh. Al_Dulamey
Hi,
You can find suitable CMOS camera.
Characteristics you mentioned may depend on the camera model
more than whether it is CCD or CMOS.
You can check your concerns based on camera datasheets.
Best,
Zoltan
CMOS cameras uses progressive scan without stopping of the exposition and therefore rolling shutter artifacts happens.
CCD cameras in most cases (or, at least, often) has global shutter, so rolling shutter do not happens.
I may assume out of "camera stabilisation tag that your experiment should measure some dynamic speckles. If so, rolling shutter may be important.
Yuriy Mironenko,Is there any way or specific settings of camera to reduce rolling shutter artifacts?something you would recommend
Muhammad Junaid Raza, I may only confirm one trick to mitigate it. Rolling shutter effect is not isotropic. It works along one axis of the frame and is negligible along the other. So if your effect is along one axis, you may just rotate your camera 90 degrees. We tried this in our latest CVPR paper and it works.
Unfortunately, I don't know any camera setting which may reduce the effect. I know a little about hardware settings of the CMOS sensors - so it's "I don't know", not "I don't think so".
Nevertheless: CMOS sensors have the ability to random access to a pixel matrix. So, if you have access to the low-level interface (i.e. programming of interaction between controller and sensor), you may alter progressive scan to something more elaborated. It's often used to scan "Regions Of Interest" only instead of the full frame, which increases frame rate. I suppose you may use same means of control to implement interlaced, double-interlaced, diagonal scan or even inward spiralling.
There are two things I should mention, however:
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Optics