• Laser & Optoelectronics Progress
  • Vol. 59, Issue 8, 0815010 (2022)
Weihua Pan, Yuanyuan Men, and Pan Su*
Author Affiliations
  • Department of Computer, North China Electric Power University, Baoding , Hebei 071000, China
  • show less
    DOI: 10.3788/LOP202259.0815010 Cite this Article Set citation alerts
    Weihua Pan, Yuanyuan Men, Pan Su. Adaptive Filtering Stereo Matching Algorithm for Fusion of Edge Features[J]. Laser & Optoelectronics Progress, 2022, 59(8): 0815010 Copy Citation Text show less
    Binary encoding process
    Fig. 1. Binary encoding process
    Local disparity maps. (a) Traditional Census combined with gradient transform algorithm; (b) proposed algorithm
    Fig. 2. Local disparity maps. (a) Traditional Census combined with gradient transform algorithm; (b) proposed algorithm
    Disparity maps obtained by different algorithms. (a) CG algorithm; (b) Cen algorithm; (c) GRD algorithm; (d) proposed algorithm
    Fig. 3. Disparity maps obtained by different algorithms. (a) CG algorithm; (b) Cen algorithm; (c) GRD algorithm; (d) proposed algorithm
    Disparity maps obtained by different algorithms. (a) True disparity map; (b) BF algorithm; (c) GF algorithm; (d) proposed algorithm
    Fig. 4. Disparity maps obtained by different algorithms. (a) True disparity map; (b) BF algorithm; (c) GF algorithm; (d) proposed algorithm
    Line chart of false matching rate
    Fig. 5. Line chart of false matching rate
    Experimental results of the proposed algorithm. (a) Left original image; (b) real disparity map; (c) unfilled disparity map; (d) final disparity map
    Fig. 6. Experimental results of the proposed algorithm. (a) Left original image; (b) real disparity map; (c) unfilled disparity map; (d) final disparity map
    AlgorithmTsukubaVenusTeddyConesAverage
    GRD2.621.617.393.543.79
    CG2.702.368.373.864.32
    Cen3.531.958.484.354.58
    Proposed algorithm2.431.327.033.323.53
    Table 1. False match rate of different algorithms
    AlgorithmTsukubaVenusTeddyConesAverage
    NoccAllDiscNoccAllDiscNoccAllDiscNoccAllDisc
    BF3.143.787.870.951.596.105.1311.8916.763.459.078.976.56
    GF2.322.626.630.450.774.544.768.7213.252.638.127.755.21
    Proposed algorithm1.852.066.450.330.543.294.057.9710.642.458.027.484.59
    Table 2. False match rate of different aggregation algorithms
    AlgorithmNoiselessSalt and pepper noise
    1%5%10%15%
    GRD3.314.486.288.269.94
    CG3.695.496.538.9310.93
    Cen4.326.287.6510.0513.94
    Proposed algorithm3.134.055.417.138.32
    Table 3. Mismatch rate under salt and pepper noise
    AlgorithmTsukubaVenusTeddyConesAverage
    GF2.815.1315.0915.569.65
    BF77.50110.74272.19271.88183.08
    ST1.962.525.565.623.91
    Proposed algorithm3.125.4515.4615.789.95
    Table 4. Running time of different algorithms
    Weihua Pan, Yuanyuan Men, Pan Su. Adaptive Filtering Stereo Matching Algorithm for Fusion of Edge Features[J]. Laser & Optoelectronics Progress, 2022, 59(8): 0815010
    Download Citation