Video Stabilization Using Modified Blurriness Index Employing De-Blurring with Temporal Derivatives Using Motion Smoothing

In low-cost portable cameras and camcorders, simple and fixed lens configurations are used.Items outside of the depth range appear fuzzy due to the short depth of focus of these cameras.Furthermore, videos taken with a hand-held camera include severe handshakes with jitters.As a result, it is critical to stabilize camera motions to increase video quality.Previous video stabilization (VS) solutions included de-blurring at the post-processing stage, resulting in erroneous measurement of motion parameters.To eliminate these estimation errors, this research suggests including de-blurring before implementing motion estimation (ME).The brightness of each pixel is convolved with the square of orthogonal temporal derivatives as a newly modified proposed blurriness index, which represents magnified objects and background motions.The proposed VS method to smooth the accidental motions and jitters in the movie has utilized an adaptive FIR filter.The simplified edge completion approach is designed to produce full-frame stabilized video sequences.The suggested method may accurately identify blurry frames at a cheap computational cost.Results present the performance comparison of the proposed blurriness index with the two most commonly used blurriness indexes.De-blurring the video before motion estimation improves the video stabilization quality.