algorithm - What are the standard techniques for removing a segmentation (such as a human or bird) from a video? -

let's taking video (with camera in steady position) , bird flies through view of camera. should possible image segmentation , automatically remove bird video.

what these styles of algorithms called , how accomplished?

there's technique called simple image object extraction (siox) - uses technique identify foreground vs. background objects in still , video images. open source gimp editor has implementation of it, , there's more information here.

from overview:

siox stands simple interactive object extraction , solution extracting foreground still images little user interaction. siox fast, noise robust, , can therefore used segmentation of videos. avoids many of drawbacks of graph-based segmentation methods performs equally on different benchmarks. siox open , free (apache license) , authors have intentionally not patented part of technology. result, has been integrated several open-source image manipulation programs on past years. siox underlying algorithm of foreground extraction tool in gnu image manipulation program (gimp) , part of tracer tool in inkscape. siox originates e-chalk instructor standing in front of electronic chalkboard segmented. variants of siox being used robotic vision , improving 3d time-of-flight camera segmentation.

here's link java reference implementation of siox.

here's link pdf details how variation of algorithm works.

you should able adapt use inter-frame interpolation remove specific foreground object each frame of video using temporal data surrounding frames.