Learn More
In this paper we evaluate and compare six well-known foreground from background subtraction methods against a standard database. To be able to compare these algorithms objectively, we have chosen three challengeable scenarios from this database. The algorithms were applied to image sequences of length 100 to 800 frames. We examined the results thoroughly(More)
ABTRACT In video surveillance, detection of moving objects from an image sequence is very important for goal tracing, action recognition, and behavior understanding. Background subtraction is a very widespread approach for foreground segmentation in a quiet still image. In order to reimburse for illumination changes, a background model apprising process is(More)
  • 1