Is there a name for oversampling a time-lapse and selecting the most similar frame each day?
Asked 3/30/2020
3 views
2 answers
0
I’m shooting a long-term garden time-lapse at about one final frame per day. To reduce flicker and jitter from changing light, clouds, wind, and other day-to-day variation, I had the idea of capturing many images each day and keeping only the single frame that best matches the previous day’s chosen frame.
My current approach compares candidate images pixel-by-pixel and selects the one with the lowest difference score. The first frame is chosen manually. It seems to work, but I’m wondering:
- Is this a known technique with an established name?
- Has it been used before for time-lapse, hyperlapse, or related work?
- Are there better ways to choose frames than a simple greedy comparison to only the previous day’s image?
I’m especially interested in terminology and any practical guidance for making the sequence smoother over long periods.
Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0
Photography Stack Exchange contributor
6y ago
2 Answers
39
Well, that is a so-called greedy algorithm that may lead into dead ends unnecessarily that are expensive to get out again. If you instead want to minimize the total change, you'd keep all images and record for each image a) the best preceding image for an optimal sequence arriving here b) the cost of arriving here via the optimal sequence.
You can prune comparisons by noting that any sequence that is already more expensive than the current optimum candidate without even adding the newest distance can be dropped without doing an actual comparison. Images that end up in "dead sequences" that can never be part of an optimum sequence regardless of what comes up next can be pruned altogether. That way, the active set of images that needs to be kept is confined, though with the close similarity of your images, there may be "close races" for a long time keeping numerous candidates active.
Originally by user91253. Source · Licensed CC BY-SA 4.0
user91253
6y ago
0
Generated from our catalog & community — verify before relying on it.
Yes—this general idea is already known, though not necessarily by one single name for your exact use case.
In photography, capturing many frames and later keeping the best ones is similar to lucky imaging, where astrophotographers take many exposures and select the least-distorted frames. More generally, oversampling and selecting later is a common technique in time-lapse, video stabilization, hyperlapse, and some action cameras’ smoothing modes.
The main caution from the answers is that choosing each day’s frame only by matching the previous selected frame is a greedy algorithm. It may make a locally good choice that causes worse jumps later. A better approach is to evaluate sequences over multiple days and minimize total transition cost, not just the immediate difference.
So: your idea is useful and related to existing techniques, but it’s not especially novel in principle. For best results, keep multiple candidate frames and optimize across a window or the full sequence rather than committing day-by-day to the single closest match.
Recommended products
UniqueBot
AI6y ago
Your Answer
Related Questions
How can I power a Canon Rebel T3i for a multi-day plant time-lapse?
What’s a simple, low-cost way to shoot an all-day time-lapse from a 5m pole outdoors?
How do I calculate the shooting interval for a 7-hour indoor time-lapse?
Do long exposures on a DSLR behave like a global shutter?
How can I create a star-trails time-lapse where previous frames linger?