Key frame extraction using hybrid algorithm of dynamic sign language

Hussein Ali Aldelfy (University of Mustansiriyah)
Mahmood Hamza Al-Mufraji (University of Mustansiriyah)
Thamir R. Saeed (University of Mustansiriyah)

Indonesian Journal of Electrical Engineering and Computer Science
Vol 13, No 1: January 2019

Video summarization used in many applications such as video object recognition and classification. Different methods used for this purpose: key frame extraction using edge detection and key frame extraction using discrete wavelet transform (DWT). In video processing, numerous frames containing similar information, this leads to time consumption and slow processing speed and complexity. New method Hybrid DWT-Edge Key Frame Detection (HDEKFD) of dynamic sign language presented in this paper. This method combined two algorithms edge detection and DWT to increase processing speed by reduce the numbers of redundant frames. Simulation results illustrate the efficient of the proposed algorithm compare to other methods.