首页 | 本学科首页   官方微博 | 高级检索  
     

曲面相机阵列多分辨成像方法
引用本文:曹杰,崔焕,孟令通,郝群,唐鸣元. 曲面相机阵列多分辨成像方法[J]. 光子学报, 2020, 49(4): 76-82. DOI: 10.3788/gzxb20204904.0411003
作者姓名:曹杰  崔焕  孟令通  郝群  唐鸣元
作者单位:北京理工大学光电学院机器人与系统教育部重点实验室 北京100081,北京理工大学光电学院机器人与系统教育部重点实验室 北京100081,中国科学院电子学研究所 北京100190,北京理工大学光电学院机器人与系统教育部重点实验室 北京100081,北京理工大学光电学院机器人与系统教育部重点实验室 北京100081
摘    要:针对现有成像系统因数据冗余而无法兼顾大视场、高分辨、高效性的问题,结合人眼视网膜变分辨成像和并列式复眼成像原理,设计一种多分辨率成像的复合仿生成像系统.该成像系统按照球面和平面兼顾的曲面布局方式,利用11个相机镜头构建相机阵列,组成了四个等级分辨率的子眼拍摄模块.通过物距100 m的远景实验和物距10 m的近景实验发现,该系统在实现高分辨成像的同时,获得总视场达150.8°×37.8°.多分辨率成像实验结果表明,该系统获取的图像的分辨率从中心视场到边缘视场逐渐降低,并且相较于中心清晰全视场成像,四级分辨率成像的拼接图像数据量减少了17.2倍的数据冗余.

关 键 词:成像系统  复合仿生成像  多分辨率  大视场  相机阵列  曲面

Multi-resolution Imaging with Camera Arrays on Curved Surface
CAO Jie,CUI Huan,MENG Ling-tong,HAO Qun,TANG Ming-yuan. Multi-resolution Imaging with Camera Arrays on Curved Surface[J]. Acta Photonica Sinica, 2020, 49(4): 76-82. DOI: 10.3788/gzxb20204904.0411003
Authors:CAO Jie  CUI Huan  MENG Ling-tong  HAO Qun  TANG Ming-yuan
Affiliation:(Key Laboratory of Biomimetic Robots and Systems,Ministry of Education,School of Optics and Photonics,Beijing Institute of Technology,Beijing 100081,China;Institute of Electrics,Chinese Academy of Sciences,Beijing 100190,China)
Abstract:Based on the variable resolution mechanism of human retina and the imaging principle of parallel compound eye,a hybrid bionic imaging system with multi-resolution is proposed to reduce data redundancy and to balance large field of view,high resolution and high efficiency.Given the consideration of both curved and plane surface on the layout,the imaging system included three layers in which the eleven cameras are divided into four groups to obtain images with four different resolutions.Two imaging experiments under the conditions of long distance of 100 m and short distance of 10 m are carried out,the results indicate that the total field of view reaches 150.8°×37.8°as keeping high resolution in the fovea.Moreover,the results of multi-resolution imaging experiments show that the system reduces the image resolution from the central field of view to the edge field of view gradually,and the amount of stitched image data with four-level resolution is reduced 17.2 times of data redundancy compared to the high-resolution image with total field of view.
Keywords:Imaging system  Hybrid bionic imaging  Multi-resolution  Large field of view  Camera array  Curved surface
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号