An alternative robust and High Reliability optical flow based on Horn-Schunck Algorithm using median filter and confidence based technique

Abstract

Optical flow is a technique for identifying the density velocity or motion vector (MV) in each pixel for motion estimation of image sequences. Under the real situation, many undesired conditions usually generate noises over the sequences and therefore this situation is corrupts the performance in reliability of optical flow. In order to increase the reliability of the density velocity, this paper proposes a novel Horn-Schucnk optical flow algorithm based on the robust estimation that using median filter and effective confidence technique using bidirectional symmetry of forward and backward flow. Comprehensive evaluations demonstrate the effectiveness results of our proposed algorithm under several standard sequences such as AKIYO, COASTGUARD, CONTAINER, and FOREMAN. These sequences have differences in foreground and background movement and speed in characteristic. These sequences are contaminated by the Additive White Gaussian Noise (AWGN) at several noise power levels (such as AWGN at 25 dB, 20 dB, and 15 dB respectively) over each sequence. Peak Signal to Noise Ratio (PSNR) is used as the performance indicator in our experiment.