Combine several images algorithms
I'm looking for algorithms that can combine images based on a quality factor. For example, you have 50-100 photographies of the same scene, but some areas had bad quality in some image because artefacts or whatever.
Now for each pixel I select the best one with a quality factor based in darkness but for sure we have a lot off possible combinations and a lot a quality measures pixel/patch/image-based.
I'm trying to research about this topic but I don't found how to describe it properly, do you know some algorithms or at least which is de name of this "problem"?
Update : Note some desired pixels or pixel areas only appears in a few cases, eg in 10 of 100 images. It causes we can't use simple averaging or similar methods.
One of the solution is averaging the images.
If you have quality factor for each sample than u can do weighted averaging.
You use following alogorithm to improve over averaging : -
Divide image into block of 4*4 or 8*8
calculate autocorrelation in all such blocks
Higher autocorrelation means lesser noise hence u can give quality factor high for autocorrealation and low otherwise.
Do weighted averaging averaging of blocks using the quality factor defined
上一篇: 在图像中查找图像（对象检测）
下一篇: 结合几种图像算法