Performance of enhanced error concealment techniques in multi-view video coding systems

Transmission of multi-view video encoded bit-streams over error-prone channels demands robust error concealment techniques. This paper studies the performance of solutions that exploit the neighbourhood spatial, temporal and inter-view information for this scope. Furthermore, different boundary distortion measurements, motion compensation refinement and temporal error concealment of Anchor frames were exploited to improve the results obtained by the basic error concealment techniques. Results show that a gain in performance is obtained with the implementation of each independent concealment technique. Furthermore, Peak Signal-to-Noise Ratio (PSNR) gains of about 4dB relative to the standard were achieved when adopting a hybrid error concealment approach.