MAP推定にもとづく再構成型超解像処理技術における画質改善

松原 史弥; 雑賀 洋平

doi:10.11511/jacc.53.0.317.0

THE 53RD JAPAN JOINT AUTOMATIC CONTROL CONFERENCE

Session ID : 644

DOI https://doi.org/10.11511/jacc.53.0.317.0

Conference information

Host: The Japan Society of Mechanical Engineers, The Society of Instrument and Control Engineers, The Institute of Systems, Control and Information Engineers, The Society of Chemical Engineers, Japan, The Japan Society for Precision Engineering, The Japan Society for Aeronautical and Space Sciences

Co-host: Technically Cosponsored by 50 Societies and Institutes

Signal Processing and Pattern Recognition

Performance Estimation of Reconstructed-Based Super-Resolution Based on Maximum A Posterior Estimate

*Fumiya Matsubara, Yohei Saika

Author information

Keywords: Super-resolution, Maximum A Posterior estimate, Prior knowledge

CONFERENCE PROCEEDINGS FREE ACCESS

Details

Abstract

We investigate noise tolerance of the Maximum A Posterior (MAP) estimate via prior knowledge on gray-level images to reconstruction-based super-resolution. With the use of numerical simulation to several standard images, we clarify that the MAP estimate accurately reconstructs an original gray-level image by removing noises introduced into a set of low-resolution corrupted images, if we assume an appropriate model of the true prior which enhances smooth structures in the pattern of gray-level images.

Corresponding author

Register with J-STAGE for free!