M
M
marshersusanin2018-09-22 19:20:43
Python
marshersusanin, 2018-09-22 19:20:43

How to improve the quality of an image pulled from a video?

Hello.
There are the following conditions: the camera shoots 1080p 30 frames per second, everything is great. During shooting, an object of interest is detected, but it is very small, which makes it very difficult to distinguish anything other than outlines.
Actually, the question arises, is it possible to improve the quality of the image in the screenshot (or its separate area)? HDR, AI, it seems, should do something similar, but they are not interested in abstractions, but in either existing solutions or algorithms that can be applied independently. Let's say programming in python.
Thanks in advance for your replies!

Answer the question

In order to leave comments, you need to log in

2 answer(s)
A
Andrey Burov, 2018-09-22
@BuriK666

No.

like in the movies
2078832_700bwp.webp

R
RobertLis, 2018-10-03
@RobertLis

If an object in the video is present on several frames of the video, then there is a chance that its resolution can be increased. Further options.
1. If the object does not change size during shooting, does not change its geometry and does not rotate, then there are free programs that will produce a clear image based on a series of frames.
2. If at the same time the object rotates, but only in the image plane, and/or changes size, then the task can be reduced to step 1 in the video editor.
3. If the object changes or rotates not in the image plane, then there are few chances to increase the resolution and there are no simple solutions.
If the object is present only on one frame, then it is impossible to reliably restore it (except for casuistic cases).

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question