• Acta Optica Sinica
  • Vol. 39, Issue 3, 0315001 (2019)
Ruihao Ma1、2、3, Feng Zhu1、3、*, Qingxiao Wu1、3, Rongrong Lu1、3, and Jingyang Wei1、3
Author Affiliations
  • 1 Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, Liaoning 110016, China
  • 2 College of Information Science and Engineering, Northeastern University, Shenyang, Liaoning 110819, China
  • 3 Key Laboratory of Opto-Electronic Information Processing, Chinese Academy of Sciences, Shenyang, Liaoning 110016, China
  • show less
    DOI: 10.3788/AOS201939.0315001 Cite this Article Set citation alerts
    Ruihao Ma, Feng Zhu, Qingxiao Wu, Rongrong Lu, Jingyang Wei. Dense Stereo Matching Algorithm Based on Image Segmentation[J]. Acta Optica Sinica, 2019, 39(3): 0315001 Copy Citation Text show less
    Flow chart of algorithm. (a) Matching cost aggregation; (b) disparity post-processing; (c) integrated flow chart
    Fig. 1. Flow chart of algorithm. (a) Matching cost aggregation; (b) disparity post-processing; (c) integrated flow chart
    Result of SLIC algorithm. (a) Tsukuba image; (b) superpixel segmentation image; (c) edge image
    Fig. 2. Result of SLIC algorithm. (a) Tsukuba image; (b) superpixel segmentation image; (c) edge image
    Results of different algorithms. (a) Hole filling; (b) cross adaptive window weighted median filtering
    Fig. 3. Results of different algorithms. (a) Hole filling; (b) cross adaptive window weighted median filtering
    Support region
    Fig. 4. Support region
    Influences of parameters on AvgPBM. (a) N; (b) β and γ; (c) ρ; (d) T
    Fig. 5. Influences of parameters on AvgPBM. (a) N; (b) β and γ; (c) ρ; (d) T
    Experimental results. (a) Left image; (b) ground-truth disparity; (c) result of proposed algorithm; (d) result of MST algorithm; (e) result of GF algorithm; (f) result of GA-DP algorithm
    Fig. 6. Experimental results. (a) Left image; (b) ground-truth disparity; (c) result of proposed algorithm; (d) result of MST algorithm; (e) result of GF algorithm; (f) result of GA-DP algorithm
    AlgorithmTsukubaVenusTeddyConesAvg PBMAvg Disc
    n-occalldiscn-occalldiscn-occalldiscn-occalldisc
    Proposed1.501.956.710.110.331.255.2710.814.52.388.027.014.997.36
    MST1.471.857.880.250.422.606.0111.614.32.878.458.105.488.08
    GF1.511.857.610.200.932.426.1611.816.02.718.247.665.558.42
    GA-DP1.572.007.320.891.003.187.2012.416.13.689.188.626.108.80
    Gray1.912.749.700.320.684.255.9911.716.23.709.6410.76.4610.21
    Table 1. AvgPBM for different algorithms%
    Stereo pairProposedGFCS-MSTGrayMST
    Tsukuba1.5021.5132.1251.9141.471
    Venus0.1110.2020.8450.3240.253
    Teddy5.2716.1657.6145.9935.532
    Cones2.3812.7124.1043.7036.015
    Alone4.5825.5344.1417.1354.633
    Art7.0819.0329.7939.88410.795
    Baby12.6214.6937.3743.2428.395
    Baby23.3016.08311.9544.91213.375
    Baby33.4615.7945.6434.5227.255
    Books8.29110.2239.56210.64510.264
    Bowling16.48114.52316.8149.77220.895
    Bowling24.8717.0839.3146.82210.155
    Cloth11.0131.0840.5111.1250.612
    Cloth22.3113.4632.8523.5744.135
    Cloth31.4612.1531.7722.2042.665
    Cloth43.2041.6221.3013.7451.873
    Dolls4.0815.0435.0026.5755.954
    Flowerpots9.80112.79216.67412.88319.415
    Lampshade15.59111.57410.4336.74211.995
    Lampshade213.88121.13520.88415.04218.203
    Laundry15.65316.40413.69218.50512.941
    Midd140.10440.11532.32236.67327.851
    Midd239.24535.85334.50236.93432.091
    Moebius7.4419.2547.6729.3358.693
    Monopoly25.40327.99522.51127.71424.212
    Plastic33.62139.29242.53440.21347.035
    Reindeer27.5747.2319.15228.3659.873
    Rocks13.6842.7022.2314.4152.833
    Rocks22.0431.6121.5712.5852.084
    Wood13.7722.8318.6844.73311.065
    Wood22.2722.3430.9913.0045.615
    AvgErr9.4210.2510.4410.7411.23
    AvgRank1.873.062.613.673.65
    Table 2. AvgPBM for different algorithms on n-occ region%
    Ruihao Ma, Feng Zhu, Qingxiao Wu, Rongrong Lu, Jingyang Wei. Dense Stereo Matching Algorithm Based on Image Segmentation[J]. Acta Optica Sinica, 2019, 39(3): 0315001
    Download Citation