Development Of Depth Video Processing Algorithm Based On Adaptive Weighted Bilateral Spatiotemporal Filter For Predicting The Visual Comfort From Stereoscopic Views

Kadmin, Ahmad Fauzan and Othman, Mohd Azlishah and Abd Rashid, Wan Norhisyam and Abd Ghani, Shamsul Fakhar and Hamzah, Rostam Affendi and Hamid, Mohd Saad (2018) Development Of Depth Video Processing Algorithm Based On Adaptive Weighted Bilateral Spatiotemporal Filter For Predicting The Visual Comfort From Stereoscopic Views. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (24 Pages) Development Of Depth Video Processing Algorithm Based On Adaptive Weighted Bilateral Spatiotemporal Filter For Predicting The Visual Comfort From Stereoscopic Views.pdf - Submitted Version Restricted to Registered users only Download (4MB)

Abstract

This work proposes an approach to capture videos with depth.That is a spatiotemporal stereo matching approach based on the adaptive weighted bilateral filter - a novel technique derived by a reformulation of an existing stereo matching approach.This is the basis for an extension which combines temporal evidence in real time, resulting in increased temporal coherence of disparity maps or depth maps,particularly in the presence of image noise.The system involves of a fourstep pipeline that aligns the video streams,matching computation,aggregated by bilateral spatiotemporal filter,capably removes and fills invalid and noisy geometry,and finally uses a median filter to increase the spatial resolution of the depth data and intensely reduce depth measurement noise. The contribution of this work is a development of a computational model that predicts the visual comfort of stereoscopic images and extending videos to comprise depth and augmenting video processing techniques.

Actions (login required)

View Item