Design and Analysis of Multi-frame Super Resolution Using OpenCV

被引：0

作者：

Liu, Tsu-Ming ^{[1
]}

Chang, Chih-Kai ^{[1
]}

Huang, Yu-Hao ^{[1
]}

Ju, Chi-Cheng ^{[1
]}

机构：

[1] Mediatek Inc, Hsinchu, Taiwan

来源：

2016 IEEE INTERNATIONAL CONFERENCE ON CONSUMER ELECTRONICS (ICCE) | 2016年

关键词：

RECONSTRUCTION; ALGORITHM; MOTION;

D O I：

暂无

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

Multi-frame super-resolution performance would highly depends on the quality of low-resolution observations. While a wide range of multi-frame super-resolution algorithms now exist, the selection design of low resolution images has not been adequately explored. This paper presents an OpenCV-based multi-frame super resolution and explores a sharpness index to optimize the low-resolution images. The experimental results reveal that the proposed frame selection design has considerable effectiveness and seventy to ninety iterations are required for convergence to desired image quality.

引用

页数：2

共 11 条

[1]

[Anonymous], P IEEE INT C IM PROC

[2] A fast super-resolution reconstruction algorithm for pure translational motion and common space-invariant blur [J].