优化设计的图像自动拼接方法资源-CSDN文库

自动图像拼接

需积分: 10 61 浏览量 2016-03-06 09:55:54 上传评论收藏 4.64MB PDF 举报

资源推荐

资源详情

资源评论

Optimized design of automatic image mosaic

Zhaoxia Fu & Liming Wang

Published online: 16 February 2013

Springer Science+Business Media New York 2013

Abstract Virtual reality technology has been widely used in the fields of aerospace, robotics

remote operation and biology medicine and so on. Panoramic image mosaic is one of the very

important parts. Since photographs taken by the ordinary camera may appear distorted, over-

lapping and tilting, we propose a wide mosaic algorithm used in the projection transformation in

this paper. The algorithm first uses the Harris operator to extract corners, adopting the improved

corner response function for avoiding the randomness of k value. Then fast RANSAC method

is used to match the images approximately, and the cross-correlation method of gray window as

the center of feature points is used to the redundant feature points for further exact match. And

then it need solve the model transformation parameters between two images according to these

corners information and obtain the projection transformation matrix. Finally, the application of

image morphing technique is for reconstructing the image having spatial transform, the result of

which are carried on stitching seamlessly with another source image. Experimental results show

that the algorithm is effective to achieve a good mosaic.

Keywords Image mosaic

Projection transformation

Image morphing

Inverse mapping

1 Introduction

The technology of image mosaic was focused of early application of remote sensing and

photography. In recent years image mosaic has gotten more and more widely used in photog-

raphy, virtual reality, video, image processing and other fields, and it has a great value. The

panoramic image obtained from stitching technique represents an effective form of the real

world, which can eliminate a lot of redundant information of image sequences and compress

storage capacity of information.

Multimed Tools Appl (2014) 72:503–514

DOI 10.1007/s11042-013-1387-y

Z. Fu (*)

L. Wang

Science and Tec hnology on Electronic Test & Measurement Laboratory and Key Laboratory of Instrumentation

Science & Dynamic Measurement(Ministr y of Education), Information and communication engineering

institute, North University of China, T aiyuan 030051, China

e-mail: fzx2005@163.com

Z. Fu

Party school of shanxi provincial committee of the C.P.C, Taiyuan 030006, China

At present, many researchers have proposed a variety of image mosaic approaches. These

reviewed approaches are mainly discussed in the frequency domain and spatial domain [5, 7,

13]. The methods of frequency domain are to use the phase correlation of the Fourier

transform and deformation technique in solving some image transformation parameters; the

methods of spatial domain are classified in area-based and feature-based algorithm according

to their nature. The area-based algorithm starts from the gray values of splicing images, and

determines the similarity by calculating the gray correlation to two images. Then the extent

and location of the overlap region of the splicing images are gotten in order to achieve image

mosaic. The feature-based mosaic algorithm searches for matching by extracting feature

points from the corresponding feature area of the overlapping par t of two images. The

algorithm has a relatively high stability and robustness, more extensive application. Image

mosaic process may encounter a variety of different situations, such as image translation,

rotation and scaling, so the applicability requirements of the mosaic algorithm also get higher

and higher. Reddy and Chatter [10] proposed a method based on Fast Fourier Transform

(FFT-based) using the polar coordinate transformation and cross-power spectrum to register

the images with translation, rotation and scaling transformation. In the method, the phase

correlation technique is simple and accurate, but it is required having a greater overlap ratio

between images. Richard Szeliski [12] proposed a projective transformation model based on

eight parameters in the two-dimensional space for image registration, using the iterative

nonlinear minimization algorithm of Levenberg-Marquardt to calculate the geometric trans-

formation parameters between images. His approach in dealing with translation, rotation,

affine transformation to splicing images has a good effect and its convergence speed is fast, so

it has become one of classic algorithms in the field of image mosaic. But in his algorithm,

extracting manually feature points to carry on estimating point transformation makes the

inadequate automation of the entire algorithm, and extracting and matching feature points

have also some inaccuracies. Matungka et al. [9] proposed a registration algorithm combining

adaptive polar transform (APT) with an innovative projection transform. The algorithm uses

the log-radius and the angular direction to express respectively the scale and rotation

parameter, which involves different scale of matching problem. The robustness and accuracy

of their algorithm are very high, but the computational complexity under the polar coordi-

nates is also very high, and the operation efficiency is not also ideal. For the shortcomings of

the existing algorithms, the purpose of this paper is to describe a common framework used in

the projection transformation within which all of these computations can be represented. The

algorithm of the paper has a stability mosaic in the different scale images, and it has also a

good robustness for translation and rotation between images.

This paper is organized as follows: Section 2 describes the whole framework of automatic

image mosaic. Section 3 gives the theory of the improved Harris operator for extracting

corners and creates the corresponding relationship of corners between images by using fast

RANSAC method and the cross-correlation method of gray window. Section 4 provides a

method of solving the projection transformation matrix. Section 5 introduces the technology

of image morphing and provides the process of image mosaic. In Section 6, we give the

experiments and analyze the experimental results. Section 7 provides the conclusions.

2 The whole framework of automatic image mosaic

The overall flowchart of the proposed algorithm is described in Fig.1. In what follows, the

details of each proposed algorithm are described. For analyzing the effectiveness of the

algorithm, the whole process of parameter estimation is given in Section 6.

504 Multimed Tools Appl (2014) 72:503–514

剩余11页未读，继续阅读

评论收藏

内容反馈

aafdf熊

粉丝: 0
资源: 9

优化设计的图像自动拼接方法

一种新的图像拼接方法

图像的自动拼接

图像拼接方法探讨

图像拼接方法的探究

基于特征点的全自动无缝图像拼接方法

OPENCV实现ORB/SURF/SIFT + RANSAC 图像自动拼接

OPENCV实现ORB/SURF/SIFT RANSAC图像自动拼接

一种改进粒子群优化算法的多机器人地图拼接方法.pdf

两种方法实现图像拼接

图像自动翘曲和拼接

MATLAB自动图像拼接

电信设备-一种基于移动终端前置摄像头的多终端屏幕自动拼接方法.zip

多幅图像拼接算法实践研究

QT + 图片拼接源码及其应用程序

基于Surf 特征配准的PCB图像拼接算法优化

基于matlab的图像拼接

全景图像自动拼接

SIFT特征的图像自动拼接.

图像分步拼接和自动拼接源码

图像拼接的多种方法的论文

图像拼接原理方法.docx

基于Levenberg—Marquardt改进算法的图像拼接

图像拼接算法

bmp和png图片拼接器

拼图--把N张图片拼接出另一张图片

图像拼接软件 PTGui.Pro.8.1.2

最新资源