Real Time Video Compression and Novelty Detection using HMM Algorithm

A new DCT approximation that possesses an extremely low arithmetic complexity, requiring only 14 additions. This novel transform was obtained by means of solving a tailored optimization problem aiming at minimizing the transform computational cost. Second, we propose hardware implementations for several 2-D 8-point approximate DCT. All introduced implementations are sought to be fully parallel time-multiplexed 2-D architectures for 8 *8 data blocks. Additionally, the proposed designs are based on successive calls of 1-D architectures taking advantage of the separability property of the 2-D DCT kernel. Designs were thoroughly assessed and compared. The proposed transform possesses low computational complexity and is compared to state-ofthe-art DCT approximations in terms of both algorithm complexity and peak signal-to-noise ratio. The proposed DCT approximation is a candidate for reconfigurable video standards such as HEVC. The proposed transform and several other DCT approximations are mapped to systolic-array digital architectures.