Acta Optica Sinica, Volume. 38, Issue 1, 0115003(2018)

RGB-D Scene Image Fusion Algorithm Based on Sparse Atom Fusion

Fan Liu*, Pengyuan Liu, Junning Zhang, and Binbin Xu
Author Affiliations
  • Shijiazhuang Campus of the Army Engineering University, Shijiazhuang, Hebei 050003, China
  • show less
    Figures & Tables(11)
    Flow chart of K-mRMR fusion algorithm
    Depth image preprocessing. (a) Original depth image; (b) image after joint bilateral filtering processing
    Scoring curves of fusion images with different atomic numbers
    (a) Dictionary atoms of original selection; (b) dictionary atoms after optimized selection
    RGB-D image fusion effect of a hat used as the identification target. (a) Depth image; (b) gray image; (c) PCA-KSVD method; (d) NSCT-KSVD method; (e) K-mRMR method
    RGB-D image fusion effect of a bowl used as the identification target. (a) Depth image; (b) gray image; (c) PCA-KSVD method; (d) NSCT-KSVD method; (e) K-mRMR method
    Joint detection of RGB-D by SIFT+SVM
    Joint detection of RGB-D based on CNN
    • Table 1. Evaluation indexes of fusion images

      View table

      Table 1. Evaluation indexes of fusion images

      AlgorithmAAVGHMIQAB/FSSIM
      PCA-KSVD5.37846.328410.63810.65040.7338
      NSCT-KSVD6.74476.476311.73890.71230.7464
      K-mRMR6.58236.884113.51730.74620.7474
    • Table 2. Comparison of recognition accuracy under different algorithms

      View table

      Table 2. Comparison of recognition accuracy under different algorithms

      AlgorithmRecognition accuracy /%
      RGB imageDepth imageRGB-D imageProposed algorithm fused image
      SIFT-SVM74.571.378.581.4
      CNN80.877.485.888.2
    • Table 3. Comparison of recognition success rate under different algorithms

      View table

      Table 3. Comparison of recognition success rate under different algorithms

      AlgorithmRecognition success rate /%
      RGB imageDepth imageRGB-D imageProposed algorithm fused image
      SIFT-SVM73.370.677.180.2
      CNN78.375.183.986.1
    Tools

    Get Citation

    Copy Citation Text

    Fan Liu, Pengyuan Liu, Junning Zhang, Binbin Xu. RGB-D Scene Image Fusion Algorithm Based on Sparse Atom Fusion[J]. Acta Optica Sinica, 2018, 38(1): 0115003

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Machine Vision

    Received: Jun. 26, 2017

    Accepted: --

    Published Online: Aug. 31, 2018

    The Author Email: Fan Liu (2434344286@qq.com)

    DOI:10.3788/AOS201838.0115003

    Topics