图像拼接算法及实现.docVIP

图像拼接算法及实现 论文关键词：图像拼接 图像配准 图像融合 全景图 论文图像拼接(iagesai)技术是将一组间重叠的图像序列空间匹配对准,经重采样合成后一幅包含各图像序列信息的宽视角场景的、完整的、高清晰的新图像的技术。图像… 论文关键词：图像拼接 图像配准 图像融合 全景图 论文图像拼接(iage sai)技术是将一组间重叠的图像序列空间匹配对准,经重采样合成后一幅包含各图像序列信息的宽视角场景的、完整的、高清晰的新图像的技术。图像拼接在摄影测量学、计算机视觉、遥感图像、医学图像分析、计算机图形学等领域有着的应用价值。 来说,图像拼接的过程由图像获取,图像配准,图像合成三组成,图像配准是整个图像拼接的基础。本文了两种图像配准算法:基于特征和基于变换域的图像配准算法。 在基于特征的配准算法的基础上,稳健的基于特征点的配准算法。改进Harris角点检测算法,所提取特征点的速度和精度。然后相似测度N(nralized rss rrelatin——归一化互),用双向最大系数匹配的方法提取出初始特征点对,用随机采样法RANSA(Rand Saple nsensus)剔除伪特征点对,特征点对的精确匹配。最后用的特征点匹配对图像的配准。本文的算法性较强,在重复性纹理、旋转角度比等较难自动匹配场合下仍可以图像配准。 Abstrat：Iage sai is a tehnlgy that arries n the spatial athing t a series f iage hih are verlapped ith eah ther, and finally builds a sealess and high quality iage hih has high reslutin and big eyesht. Iage sai has idely appliatins in the fields f phtgraetry, puter visin, rete sensing iage pressing, edial iage analysis, puter graphi and s n. 。In general, the press f iage sai by the iage aquisitin, iage registratin, iage synthesis f three steps, ne f iage registratin are the basis f the entire iage sai. In this paper, t iage registratin algrith: Based n the harateristis and transfr dain-based iage registratin algrith. In feature-based registratin algrith based n a rbust feature-based registratin algrith pints. First f all, t iprve the Harris rner detetin algrith, effetively iprve the extratin f feature pints f the speed and auray. And the use f a siilar easure f N (nralized rss rrelatin - Nralized rss-rrelatin), thrugh the largest rrelatin effiient ith t-ay athing t extrat the feature pints ut the initial right, using rand sapling ethd RANSA (Rand Saple nsensus) exluding pseud-feature pints right, feature pints n the ipleentatin f the exat ath. Finally ith the rret feature pint athing fr iage registratin ipleentatin. In this paper, the algrith adapted, in the repetitive texture, suh as relatively large rtatin re diffiult t autatially ath asins an still ahieve an aurate iage registr