拍摄全景建筑图像自动拼接匹配仿真
|
摘要
对全景建筑不同角度图像的自动拼接匹配,能够有效提高对建筑图像的复原。对全景图像的自动拼接匹配,需要提取全景图像特征区域,对匹配块重叠区域链接约束进行局部优化,完成全景图像的拼接匹配。传统方法获得待拼接图像的显著图,将待拼接图像的前景和背景区域分离,但忽略了优化局部区域链接约束,导致拼接效果不理想。提出基于局部边缘密度LED算法的全景图像自动拼接匹配方法。结合视觉传达技术进行全景图像特征区域提取,采用待拼接图像的局部熵差进行特征点匹配搜索,对特征匹配块重叠区域引入形状扭曲约束和块链接约束进行局部优化,获得修正后的待拼接图像,对待配准图像进行图像融合获得全景拼接图像。实验结果表明,所提方法能够有效突出图像边缘细节,实现了尺度变化条件下的全景图像无缝拼接。
Traditional method ignores the optimization of link constraint in local area,resulting in the unideal splicing effect. An automatic stitching matching method of panoramic image of LED algorithm based on local edge density is proposed. Combined with visual communication technology,feature region of panoramic image is extracted.The local entropy difference of image to be spliced is used for matching search of feature point. The shape warping constraint and block chaining constraint are introduced into overlap region of feature matching block for local optimization,so as to obtain modified images which will be spliced. Finally,image composite is used for registering images to obtain panoramic splicing image. Simulation results show that the proposed method can effectively emphasize edge details of images,which achieves seamless stitching of panoramic image at scale variation.
Traditional method ignores the optimization of link constraint in local area,resulting in the unideal splicing effect. An automatic stitching matching method of panoramic image of LED algorithm based on local edge density is proposed. Combined with visual communication technology,feature region of panoramic image is extracted.The local entropy difference of image to be spliced is used for matching search of feature point. The shape warping constraint and block chaining constraint are introduced into overlap region of feature matching block for local optimization,so as to obtain modified images which will be spliced. Finally,image composite is used for registering images to obtain panoramic splicing image. Simulation results show that the proposed method can effectively emphasize edge details of images,which achieves seamless stitching of panoramic image at scale variation.
引文
[1]朱宁宁.柱面全景图像拼接方法的仿真分析[J].测绘学报,2017,46(4):487-497.
[2]侯天怡,尤红建,孙显.以卫星影像为底图的无人机图像自动拼接方法[J].测绘科学,2015,40(6):11-16.
[3]秦绪佳,等.基于最佳缝合线的序列遥感图像拼接融合方法[J].计算机科学,2015,42(10):306-310.
[4]蒋波,翟旭平.基于PCA-SIFT特征匹配的图像拼接算法[J].计算机应用,2016,36(s2):143-145.
[5]董强,等.基于改进BRISK的图像拼接算法[J].电子与信息学报,2017,39(2):444-450.
[6]樊晓婷,等.信息熵和差分激励融合的图像拼接检测[J].计算机工程与设计,2016,37(1):37-41.
[7]韩智慧,赵建华.一种用于DBN图像局部变形方法的改进研究仿真[J].计算机仿真,2016,33(12):305-308.
[8]王丹丹,陈康.基于虚拟现实的柱面全景图像拼接方法研究.[J].科技通报,2017,10(33):113-116.
[9]张勇,王志锋,马忍.基于改进SIFT和Contourlet变换的图像拼接方法研究[J].微电子学与计算机,2016,33(10):134-138.
[10]王凯,等.一种改进的非线性加权图像拼接融合方法[J].小型微型计算机系统,2017,38(5):1116-1120.
[2]侯天怡,尤红建,孙显.以卫星影像为底图的无人机图像自动拼接方法[J].测绘科学,2015,40(6):11-16.
[3]秦绪佳,等.基于最佳缝合线的序列遥感图像拼接融合方法[J].计算机科学,2015,42(10):306-310.
[4]蒋波,翟旭平.基于PCA-SIFT特征匹配的图像拼接算法[J].计算机应用,2016,36(s2):143-145.
[5]董强,等.基于改进BRISK的图像拼接算法[J].电子与信息学报,2017,39(2):444-450.
[6]樊晓婷,等.信息熵和差分激励融合的图像拼接检测[J].计算机工程与设计,2016,37(1):37-41.
[7]韩智慧,赵建华.一种用于DBN图像局部变形方法的改进研究仿真[J].计算机仿真,2016,33(12):305-308.
[8]王丹丹,陈康.基于虚拟现实的柱面全景图像拼接方法研究.[J].科技通报,2017,10(33):113-116.
[9]张勇,王志锋,马忍.基于改进SIFT和Contourlet变换的图像拼接方法研究[J].微电子学与计算机,2016,33(10):134-138.
[10]王凯,等.一种改进的非线性加权图像拼接融合方法[J].小型微型计算机系统,2017,38(5):1116-1120.