Why do dark circular patches appear in CCTV night images after headlights pass through the frame?
Asked 2/9/2015
1 views
2 answers
0
I'm capturing JPEG stills from an Axis P1357-E IP CCTV camera at night and seeing circular dark patches on the road in places that were brightly lit by a vehicle's headlights in the previous moments. The effect shows up across consecutive frames and seems evenly spaced, suggesting it follows the motion of the headlights. The camera uses a small progressive-scan CMOS sensor, with constant IR illumination on the scene. Available settings include exposure value, exposure control, shutter speed, gain, and consecutive frames. What causes this artifact, and are there camera settings that might reduce it?
Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0
Photography Stack Exchange contributor
11y ago
2 Answers
1
I posit this as a possible explanation:
Since the spacing of the effect is uniform, this suggests that the latent images you see are the electronic remnants of 'hottest' parts of the previous few frames that were captured by the CMOS sensor. I'm not suggesting that this is a collage of multiple captures, but that the sensor has an inherent 'refresh rate' of sorts, and that the electrical signal from previous refreshes takes some time to fade to zero, so any pixels with a remnant signal >0 may still be visible at any given time - i.e. the whitest parts of the previous images. As for why the whitest parts of previous frames appear as black (e.g. black headlight circles), I can only guess that this is also part of the processor design, but I'm sure some electrical engineers out there can explain this behaviour better.
If this assumption is correct the effect would not be visible during the daytime, as the local contrast of the image would be much lower.
An analogy would be to look at consumer LCD televisions displaying high contrast motion: e.g. a white ball moving rapidly across a black background. Some televisions (depending on panel type) will show a very noticeable latent image of the white ball trailing behind the current frame. This happens because it takes X milliseconds for the LCD panel to fade white pixels back to black. The longer this fade takes, the more noticeable the effect.
Originally by user38928. Source · Licensed CC BY-SA 4.0
user38928
11y ago
0
Generated from our catalog & community — verify before relying on it.
The most likely cause is image persistence/lag from the CMOS sensor or its processing pipeline after very bright highlights hit the sensor. In low light, headlights can strongly overdrive parts of the image; remnants of that signal may persist into the next few frames. Because the bright areas were saturated, the residual can show up as dark patches after processing rather than bright ones.
The even spacing across consecutive frames supports this: each patch likely corresponds to where the headlights were in earlier frames.
To reduce it, try limiting highlight overload rather than treating it as a JPEG issue:
- use a faster shutter speed if possible
- reduce exposure/gain so headlights clip less severely
- avoid overly aggressive exposure settings in very dark scenes
In short, this is probably a sensor/processing artifact from intense moving lights, not something unique to MJPEG/JPEG still extraction. Lowering exposure and gain, and shortening exposure time, are the most likely settings changes to help.
Recommended products
UniqueBot
AI11y ago
Your Answer
Related Questions
Why does dark-frame subtraction add noise and shift color in my night photos?
How can I minimize the gap between 8-second exposures on a Canon 200D?
When should I shoot dark frames for long tracked astrophotography exposures?
Why can aperture change between nearly identical shots in shutter-priority mode?
Why do long-exposure dark frames show bands, colored patches, and bright areas?