简化变换模型的航拍图像拼接技术  

The Aerial Image Mosaic Technology Based on Simplified-transformation Model

作　　者：李艳玲[1] 汪国平[2] LI Yanling;WANG Guoping(Department of Computer,Changzhi University,Shanxi Changzhi 046011,China;School of Electronics Engineering and Computer Science,Peking University,Beijing 100871,China)

机构地区：[1]长治学院计算机系,山西长治046011 [2]北京大学信息科学技术学院,北京100871

出　　处：《弹箭与制导学报》2021年第4期5-8,共4页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：国家自然科学基金(61472010);国家重点研发计划(2017YFB1002700);山西省基础研究计划(201601D011007);山西省高等学校科技创新项目(2019L0911);长治学院校级科研项目(GJZXM201902)资助。

摘　　要：航拍图像已广泛应用于虚拟现实技术等领域,但单次拍摄难以取得较好的效果,需要将多张图像拼接起来。针对多视点航拍图像拼接范围受限、融合开销大的缺点,从简化变换模型和局部图像融合两个方面探讨提高图像拼接效果的途径。平面场景假设下,相似变换模型近似取代图像间映射,可在保留原图像视点和拼接精度之间取得平衡。结合分块曝光补偿做局部图像融合,降低时间与空间开销。实验表明,该方法可以扩大拼接视角,减少光照差异较大图像间的亮度差异。Aerial image has been widely used in virtual reality technology and other fields.However,it is difficult to achieve good results in one shot,so image mosaic technology is needed.To solve the problem of multi-view remote sensing image mosaic,we discuss the ways to improve the effect of image mosaic from two aspects of simplified transformation model and image fusion.Under the assumption of planar scene,the similarity transformation model approximates image mapping to achieve a balance between preserving the original image view point and stitching accuracy.Combined with block exposure compensation,local image fusion can reduce time and space overhead.Experimental results show that this method can enlarge the field of mosaic and reduce the brightness difference between images.

关 键 词：图像拼接 相似变换 曝光补偿 图像融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]