高级检索

微透镜阵列实现3维物体旋转不变实时识别

郝劲波, 王良甚, 忽满利

郝劲波, 王良甚, 忽满利. 微透镜阵列实现3维物体旋转不变实时识别[J]. 激光技术, 2009, 33(1): 8-11,41.
引用本文: 郝劲波, 王良甚, 忽满利. 微透镜阵列实现3维物体旋转不变实时识别[J]. 激光技术, 2009, 33(1): 8-11,41.
HAO Jin-bo, WANG Liang-shen, HU Man-li. Real-time 3-D object rotation-invariant recognition with microlens array[J]. LASER TECHNOLOGY, 2009, 33(1): 8-11,41.
Citation: HAO Jin-bo, WANG Liang-shen, HU Man-li. Real-time 3-D object rotation-invariant recognition with microlens array[J]. LASER TECHNOLOGY, 2009, 33(1): 8-11,41.

微透镜阵列实现3维物体旋转不变实时识别

详细信息
    作者简介:

    郝劲波(1979- ),男,讲师,主要从事光信息处理和光学模式识别研究.E-mail:haojinbo@sina.com

  • 中图分类号: O438.2

Real-time 3-D object rotation-invariant recognition with microlens array

  • 摘要: 为了实现3维物体旋转不变实时识别,应用微透镜阵列的多视角成像特点,利用透射像阵列的高关联性,实现3维物体信息与2维透射像阵列信息之间的转换,从而可以利用光学2维图像识别技术实现3维物体的识别。对转换和识别过程进行了理论分析,用匹配滤波的方法进行了实验验证,实现了3维物体旋转不变实时识别。得到了良好的识别效果,并实现了旋转方向的准确定位和旋转角度大小的比较判别。结果表明,应用微透镜阵列可以实现旋转3维物体旋转不变实时识别。
    Abstract: In order to realize 3-D object rotation-invariant recognition, with help of high relevancy of perspective array of 3-D objects, the depth information of 3-D objects was converted to 2-D elemental images by means of microlens array.Theoretical analysis showed that both 3-D object rotation-invariant recognition and exact orientation of rotation direction could be obtained.Taking a dice,a 3-D object as an example, it was recognized in real time by means of match filtering and the proposed method.The latter performed better, and the exact orientation of rotation direction and rotational angle size discrimination were recognized.The results showed that the real-time 3-D object rotation-invariant recognition can be accomplished with microlens array.
  • [1]

    SONG F J,JUTAMULIA S.Advanced optical information processing[M].Beijing:Peking University Press,1998:64-77(in Chinese).

    [2]

    SHENG Z X,SUN X L.Application of computer generated hologram technology in optical correlation recognition system[J].Laser Technology,2008,32(1):61-63(in Chinese).

    [3]

    CAO F,SU X Y.Method for 3-D objects recognition based on phaseencoded range image[J].Laser Journal,2007,28(1):50-52(in Chinese).

    [4]

    LU J,SU XY,WANGH X.Three dimensional object recognition with rotation-invariant by structure light[J].Journal of Optoelectronics·Laser,2004,15(12):1492-1497(in Chinese).

    [5]

    WANG H X,SU X Y,LU J.Method for 3-D ohject recognition by Moire fringe[J].Journal of Optoelectronics·Laser,2005,16(3):349-353(in Chinese).

    [6]

    LI Q,ZHOU M L,LIU J.Multi-resolution mesh based 3-D object recognitiou[J].Acta Electmnica Sinica,2001,29(7):891-894(in Chinese).

    [7]

    JAVIDI B,TAJAHUERCE E,MARTINEZ-CORBAL M,et al.New developments in active and passive 3-D image sensing,visualization,and processing[J].SPIE,2005,5986:598601.1-598601.11

    [8]

    NELLERIA,GOPINATHANU,JOSEPH J,et al.Three-dimensional object recognition from digital Fresnel hologram by wavelet matched filtering[J].Opt Commun,2006,259(2):499-506.

    [9]

    MATOBA O,TAJAHUERCE E.JAVIDI B.Real-time three-dimension object recognition with multiple perspectives imaging[J].Appi Opt,2001,40(20):3318-3325.

    [10]

    HONG S H,JANG J S,JAVIDI B.Three-dimensional volumetric object reconstruction using computational integral imaging[J].Optics Express,2004,12(3):483-491.

    [11]

    FRAUEL Y,TAJAHUERCE E,MATOBA O,et al.Comparison of passive ranging integral imaging an active imaging digital holography for three-dimensional object recognition[J].Appl Opt,2004,43(2):452-462.

计量
  • 文章访问数:  2
  • HTML全文浏览量:  0
  • PDF下载量:  12
  • 被引次数: 0
出版历程
  • 收稿日期:  2007-10-17
  • 修回日期:  2008-01-06
  • 发布日期:  2009-02-24

目录

    /

    返回文章
    返回