Multi-visual Tasks Based on Night-Vision Data Structure and Feature Analysis

Data structure and feature analysis is the first step which provides inputs of many computer vision algorithms. Thus, it is vitally important to ensure its speed, accuracy and robustness. Currently, there is a wide performance gap between the slow feature analysis and the much higher requests for fa...

Full description

Saved in:
Bibliographic Details
Published inNight Vision Processing and Understanding pp. 45 - 85
Main Authors Han, Jing, Yue, Jiang, Bai, Lianfa
Format Book Chapter
LanguageEnglish
Published Singapore Springer Singapore Pte. Limited 2019
Springer Singapore
Online AccessGet full text
ISBN9789811316685
9811316686
DOI10.1007/978-981-13-1669-2_3

Cover

More Information
Summary:Data structure and feature analysis is the first step which provides inputs of many computer vision algorithms. Thus, it is vitally important to ensure its speed, accuracy and robustness. Currently, there is a wide performance gap between the slow feature analysis and the much higher requests for faster real-time solutions. This chapter introduces the active research field of visual feature analysis, referring to literature review, addressing the detail techniques and demonstrating how to apply these features in some vision tasks. Such methods utilise low-level processing to determine valuable vision tasks, while using feature attributes, such as super-resolution (SR), superpixel segmentation and saliency.
ISBN:9789811316685
9811316686
DOI:10.1007/978-981-13-1669-2_3