计算机科学 ›› 2017, Vol. 44 ›› Issue (4): 302-305.doi: 10.11896/j.issn.1002-137X.2017.04.061

• 图形图像与模式识别 • 上一篇    下一篇

3D车载环视全景生成方法

刘冬,秦瑞,陈曦,李庆   

  1. 中国科学院微电子研究所昆山分所 昆山215347;中国科学院物联网研究发展中心 无锡214135,中国科学院微电子研究所昆山分所 昆山215347;中国科学院微电子研究所 北京100871,中国科学院微电子研究所昆山分所 昆山215347;中国科学院物联网研究发展中心 无锡214135,中国科学院微电子研究所 北京100871
  • 出版日期:2018-11-13 发布日期:2018-11-13
  • 基金资助:
    本文受移动物联网关键技术与应用(XDA06040300),汽车移动物联网总体技术研究(2012ZX03005018),面向智慧交通的车联网高安全性内容保障技术研究与核心产品研发(2012B0910)资助

Generation of Three-dimention Vehicle Panorama

LIU Dong, QIN Rui, CHEN Xi and LI Qing   

  • Online:2018-11-13 Published:2018-11-13

摘要: 通过单应矩阵生成的鸟瞰视角全景图像存在严重的信息损失和边缘模糊形变。为三维空间点加入约束,将其限制在二维空间中,可以实现一幅图像从一个视角唯一变换另一视角。假设3D空间点来自于一个“包裹”车辆的曲面,使得原始摄像头图像上每一点对应的投影线都和曲面有交点,摄像头从不同角度观察曲面和车辆,便可生成多视角全景图像。针对相交区域也进行了特殊处理。最后的实验表明,生成的多视角的全景图像既可以充分利用原始图像信息从不同视角观测车辆周边场景,同时也减少了边缘模糊和扭曲,相交区域的过渡也显得更加自然、平滑。

关键词: 车载环视系统,曲面约束,多视角全景

Abstract: Information loss,image blur and object deformation exist in bird-eye panoramic image generated by homography matrix.Confined 3D point to two-dimensional space,image in a view can transform to another uniquely.We selected a curved surface wrapping the vehicle as selected two-dimensional space.We can get the multi-view panorama image from different perspectives.A special process is also applied to the overlapping region.The final experiment demonstrate that,the multi-view panorama images make use of all the information in the source images,at the same time,image blur and object deformation decreases notably.The overlapping regions appear more naturally and smoothly.

Key words: Vehicle around view system,Surface constraint,Multi-view panorama

[1] KATO K,SUZKUK M,FUJITA Y,et al.Image synthesis display method and apparatus for vehicle camera: U.S.Patent 7,9,412[P].2006.
[2] BAO Q Q.Research on algorithm of vehicle panoramic system based on projection to variable curved surface[D].Qinghuangdao:Yanshan University,2010.(in Chinese) 鲍千千.基于可变曲面投影的汽车全景系统算法研究[D].秦皇岛:燕山大学,2010.
[3] TANIGUCHI S,YAMADA M,HAYASHIDA T,et al.MultiAngle Vision System to Supplement Driver’s Visual Field[C]∥ 17th ITS World Congress.2010.
[4] SCARAMUZZA D,MARTINELLI A,SIEGWART R.A flexible technique for accurate omnidirectional camera calibration and structure from motion[C]∥IEEE International Conference on Computer Vision Systems,2006(ICVS’06).IEEE,2006:45-45.
[5] CHANG J Y,QIN R,LI Q,et al.Image Quality Assessment of Panoramic Imag[J].Computer Science,2014,41(6):278-281.(in Chinese) 常嘉义,秦瑞,李庆,等.全景鸟瞰拼接图像的质量评价方法[J].计算机科学,2014,41(6):278-281.
[6] SUHR J K,JUNG H G.Sensor Fusion-Based Vacant Parking Slot Detection and Tracking[J].IEEE Transactions on Intelligent Transportation Systems,2014,15(1):21-36.
[7] GU X L.Panorama-based detection and recognition of road ma-rkings[D].Shanghai:Shanghai Jiao Tong University,2014.(in Chinese) 顾霄琳.基于全景环视的道路交通标线检测与识别方法研究[D].上海:上海交通大学,2014.
[8] 陈秀新,邢素霞.图像/视频检索与图像融合[M].北京:机械工业出版社.2012:130-224.
[9] BAGGIO D L,EMAMI S,ESCRIV D M,et al.Mastering Open-CV with practical computer vision projects[M].Birmingham:Packt Publisbishing.2012.
[10] CHOI Y W,KIM K Y,PARK M R,et al.Development of around view system based on moving object detection for safe bus getting on and off[C]∥ 2015 IEEE International Symposium on Consumer Electronics (ISCE).IEEE,2015.
[11] JEON B,PARK G,LEE J,et al.A Memory-Efficient Architecture of Full HD Around View Monitor Systems[J].IEEE Transactions on Intelligent Transportation Systems,2014,15(6):2683-2695.

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!