Active Mesh Based Motion Compensation Algorithm for Video Compression in Wavelet Sub Bands
Vidyasagar K N1, Bharathi S H2, Vinay N A3
1Vidyasagar K N, School of ECE, REVA Universty, Bengaluru, India.
2Bharathi S H, School of ECE, REVA University, Bengaluru, India.
3Vinay N A, School of ECE, REVA University, Bengaluru, India.
Manuscript received on July 20, 2019. | Revised Manuscript received on August 10, 2019. | Manuscript published on August 30, 2019. | PP: 1812-1817 | Volume-8 Issue-6, August 2019. | Retrieval Number: E7171068519/2019©BEIESP| DOI: 10.35940/ijeat.E7171.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The method of assessing the motion vectors is termed as Motion Estimation and consequently reconstructing the original frame using these motion vectors is termed as Motion Compensation. In this paper a new active mesh based Motion estimation and compensation algorithm in wavelet domain at subpixel level is proposed. To achieve the accuracy in estimating the motion vector from reference frame to predicted frame in sub pixel level a Wavelet based Counterlet transform is used. This new mesh based method is based by minimizing the mesh energy with energy functions as its input coefficients. The accuracy of the estimation and the compensation is improved by the feature points obtained by the energy functions. The analysis of the experimental results describes that there is improvement in the distortion of the performance rate for the video compression and this is achieved by the generation of the structured and the unstructured active mesh motion techniques in the wavelet sub-bands. The energy criteria of the input coefficients helps to retain the energy in the sub-bands which is a condition that helps in significant gaining of the PSNR required for better outcome.
Keywords: Motion Estimation, Motion Estimation, Active mesh.