Motion History Images

From HandWiki

The motion history image (MHI) is a static image template helps in understanding the motion location and path as it progresses.[1] In MHI, the temporal motion information is collapsed into a single image template where intensity is a function of recency of motion. Thus, the MHI pixel intensity is a function of the motion history at that location, where brighter values correspond to a more recent motion. Using MHI, moving parts of a video sequence can be engraved with a single image, from where one can predict the motion flow as well as the moving parts of the video action.

Some important features of the MHI representation are:[2]

  • It represents motion sequence in a compact manner. In this case, the silhouette sequence is condensed into a grayscale image, where dominant motion information is preserved.
  • MHI can be created and implemented in low illumination conditions where the structure cannot be easily detected otherwise.
  • The MHI representation is not so sensitive to silhouette noises, holes, shadows, and missing parts.
  • The gray-scale MHI is sensitive to the direction of motion because it can demonstrate the flow direction of the motion.
  • It keeps a history of temporal changes at each pixel location, which then decays over time.
  • The MHI expresses the motion flow or sequence by using the intensity of every pixel in a temporal manner.

General algorithm

for each time t
    Bt := absolute_difference(It, It-1) > threshold
end for

for each time t
    for each pixel (x, y)
        if Bt(x, y) = 1
            MHIt(x, y) := τ
        else if MHIt-1 ≠ 0
            MHIt(x, y) := MHIt-1(x, y) - 1
        else
            MHIt(x, y) := 0
        end if
end for

References

  1. "Research: Motion History Images". http://web.cse.ohio-state.edu/CVL/Research/MHI/mhi.html. Retrieved 13 November 2014. 
  2. Ahad, Md Atiqur Rahman. Motion history images for action recognition and understanding. Springer Science & Business Media, 2012.