Bilateral video super-resolution using non-local means with adaptive parameters

Super-resolution (SR) algorithms for video sequences with high resolution (HR) guide frames can provide outstanding performances. Non-local means (NLM) algorithm compares the similarity between a pixel and its neighbors. NLM replaces every pixel with a weighted average of its neighbors. The NLM base...

Full description

Saved in:
Bibliographic Details
Published in2016 IEEE International Conference on Image Processing (ICIP) pp. 1155 - 1159
Main Authors Yawei Li, Xiaofeng Li, Zhizhong Fu, Xiuxia Yin, Yufei Zhao
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2016
Subjects
adaptive parameters
Hafnium
Image reconstruction
Interpolation
Motion estimation
non-local means
Spatial resolution
Super-resolution
video scaling
Video sequences
Online AccessGet full text

Cover

More Information
Summary:Super-resolution (SR) algorithms for video sequences with high resolution (HR) guide frames can provide outstanding performances. Non-local means (NLM) algorithm compares the similarity between a pixel and its neighbors. NLM replaces every pixel with a weighted average of its neighbors. The NLM based SR algorithm can super-resolve low resolution (LR) frames using the HR guide frames in the video sequence. However, the fixed decaying factor of NLM cannot satisfy regions of distinct characteristics in a LR frame. The fixed searching window fails to balance the requirements of low computational complexity and good SR images. Thus, we propose novel criteria for selecting the decaying factor and searching window adaptively. Bilateral adjacent HR frames are used to handle the occlusion problem. The experimental results verify the validity of the proposed method.
ISSN:2381-8549
DOI:10.1109/ICIP.2016.7532539