首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于Gabor小波模型图象配准的计算视觉新方法
引用本文:杨敬安.基于Gabor小波模型图象配准的计算视觉新方法[J].合肥工业大学学报(自然科学版),1999(1).
作者姓名:杨敬安
作者单位:合肥工业大学计算机与信息学院
基金项目:国家自然科学基金高技术项目
摘    要:文章提出由两幅部分重叠图象估计物体平移、旋转和缩放新的计算视觉方法。该方法有以下特点:利用照明方向估计算子以获得摄象机旋转的起始估计;为检测局部表面曲率间断,利用Gabor小波模型定位特征点;平移及旋转的估计可公式化为线性问题;利用由粗到细的层级匹配能获得很好的结果。最后给出该算法的某些应用

关 键 词:摄象机运动模型  Gabor小波模型  立体图象配准  特征点提取

A COMPUTATIONAL VISION APPROACH TO MOTION ESTIMATION BASED ON TWO PARTIALLY OVERLAPPING IMAGES
Yang Jing''''an.A COMPUTATIONAL VISION APPROACH TO MOTION ESTIMATION BASED ON TWO PARTIALLY OVERLAPPING IMAGES[J].Journal of Hefei University of Technology(Natural Science),1999(1).
Authors:Yang Jing'an
Institution:Hefei University of Technology
Abstract:Automatic image registration is important for many multiframe based image analysis applications.In this paper,computational vision approach is presented for estimation of 2D translation,rotation,and scale from two partially overlapping images.The approach has the following features: illuminant direction estimator is used to obtain an initial estimate of camera rotation, feature points are located based on a Gabor wavelet model for detecting local curvature discontinuities, estimation of rotation and translation is formulated as a linear problem and hierarchical coarse to fine matching is ued.This results in a fast and novel algorithm that produces good results even when large rotations and scale changes have occurred between the two frames and the images are devoid of significant features.Several applications of the algorithm are presented.
Keywords:camera motion model  Gabor wavelet model  stereo image registration  feature point extraction
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号