基于无人机遥感视频影像的拼接缝消除算法  

Seam Removal Based on the Unmanned Aerial Vehicle Remote Sensing Mosaic Images

作　　者：胡惠平[1] 陈水利[2] 吴云东[2] 

机构地区：[1]福州大学数学与计算机科学学院,福建福州350000 [2]集美大学理学院,福建厦门361021

出　　处：《江南大学学报（自然科学版）》2013年第6期672-676,共5页Joural of Jiangnan University (Natural Science Edition)　

基　　金：福建省自然科学基金项目(2011J01013);福建省科技厅产学研重大项目(2011H6020);厦门市科技计划项目(3502Z20123022)

摘　　要：针对多时影像存在色调和亮度差异、变换参数不精确等因素,导致拼接后的图像存在明显的拼接缝,传统的拼接缝消除方法(渐入渐出融合,以及动态宽度等算法)处理后的影像无法完全消除垂直或水平拼接缝,提出了一种新的拼接缝消除方法。将复杂的多度重叠区域问题简化为两度重叠区域,使多条垂直和水平拼接缝减少到一条。通过一组航拍视频图像的仿真实验证明该拼接缝消除算法提高了拼接后的视觉效果。Multi-temporal images have different grayscales and textures. The inaccurate transformation parameters may result in seam in mosaic images. Traditional algorithms, such as the incremental-based strategies and dynamic-based methods will result in multi-seam in the vertical and horizontal directions. Motivated by this, this paper presents a new seam removal algorithm, which aims at transforming the complex overlap problem to double overlap and at reducing the seams from vertical or horizontal direction. The experimental results obtained from the unmanned aerial vehicle video images show that the proposed method has a better vision appearance, compare with the traditional methods.

关 键 词：拼接缝消除 渐入渐出融合算法 动态宽度拼接缝消除算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]