A new speed-based adaptive motion estimation algorithm for video sequences
Sherin Abdelaal, M. B. Abdelhalim, Abdelfattah A. Hegazy · 2014
The paper proposed a new motion speed-based adaptive algorithm (MSBAA), it is adaptive to motion speed variations in frames that occur during real time video streaming. The new proposed algorithm selects certain algorithms to be applied on the tested frames according to the type of motion detected in these frames. So in case fast motion is detected in the tested frames, the algorithm selects the fast adaptive rood pattern search algorithm ARPS to be applied on the frames. And in case slow motion is detected in the frames the new motion speed-based adaptive algorithm (MSBAA) selects a new proposed slow motion hybrid full search FS&ARPS algorithm to be applied on the frames, this mechanism is done continuously until ending the real time video streaming. The new MSBAA algorithm is adaptive as it adapts itself to the type of motion in the video scenes, and chooses the most suitable algorithm for each type of motion it detects.