基于梯度信息的新颖多源图像快速配准方法

Novel quick registration method of multisource image based on gradient information

Abstract

本发明涉及基于梯度信息的新颖多源图像快速配准方法,属于图像处理技术领域。本发明实现包括第一步特征点检测和定位,第二步主方向确定,第三步描述符生成,第四步特征点匹配。
The invention relates to a novel quick registration method of a multisource image based on gradient information, and belongs to the field of image processing technologies. The implementation of the novel quick registration method comprises a first step of detecting and locating feature points, a second step of determining a principal direction, a third step of generating a descriptor, and a fourth step of matching the feature points.

Claims

Description

Topics

Download Full PDF Version (Non-Commercial Use)

Patent Citations (1)

    Publication numberPublication dateAssigneeTitle
    CN-102521837-AJune 27, 2012湖南工业大学基于全局信息的图像配准方法

NO-Patent Citations (3)

    Title
    DONG ZHAO 等: "Rapid multimodality registration based on MM-SURF", 《HTTP://WWW.SCIENCEDIRECT.COM/SCIENCE/ARTICLE/PII/S0925231213011247》, 22 November 2013 (2013-11-22), XP028614814, DOI: doi:10.1016/j.neucom.2013.10.037
    温江涛等: "基于局部灰度梯度特征的图像快速配准方法", 《清华大学学报(自然科学版)》, vol. 49, no. 5, 31 December 2009 (2009-12-31), pages 673 - 675
    王君本等: "一种基于快速鲁棒特征的图像匹配算法", 《计算机工程与科学》, vol. 33, no. 2, 31 December 2011 (2011-12-31), pages 112 - 117

Cited By (3)

    Publication numberPublication dateAssigneeTitle
    CN-104036480-BMarch 08, 2017天津大学基于surf算法的快速消除误匹配点方法
    CN-104240231-ADecember 24, 2014大连民族学院基于局部结构二进制模式的多源图像配准
    CN-104978709-AOctober 14, 2015北京邮电大学描述符生成方法及装置