Semi-Global Stereo Matching of Remote Sensing Images Combined with Speeded up Robust Features

Yangping Wang; Anna Qin; Qi Hao; Jianwu Dang

doi:10.3788/AOS202040.1628003

Acta Optica Sinica, Volume. 40, Issue 16, 1628003(2020)

Semi-Global Stereo Matching of Remote Sensing Images Combined with Speeded up Robust Features

Yangping Wang^1,2, Anna Qin^1、*, Qi Hao³, and Jianwu Dang^1,2

Author Affiliations

¹School of Electronic and Information Engineering, Lanzhou Jiaotong University, Lanzhou, Gansu 730070, China

²Gansu Engineering Research Center for Artificial Intelligence and Graphic & Image Processing, Lanzhou, Gansu 730070, China;

³Xi′an Aerospace Data Technology Co., Ltd., Xi′an, Shaanxi 710100, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(9)

Fig. 1. Flow chart of proposed algorithm

Download full size

Fig. 2. Diagram of Census conversion

Download full size

Fig. 3. Comparison of disparity refinement algorithms. (a) Algorithm in Ref.[18]; (b) proposed algorithm

Download full size

Fig. 4. Experiment results in WorldView stereo pair. (a) Left image; (b) right image; (c) feature point extraction in left image; (d) feature point extraction in right image; (e) disparity map of SGM algorithm based on mutual information; (f) disparity map of proposed algorithm; (g) disparity map of WACS algorithm; (h) rendering disparity map of proposed algorithm; (i) refined disparity map

Download full size

Fig. 5. Experimental results in IKONOS stereo pair. (a) Left image; (b) right image; (c) feature point extraction in left image; (d) feature point extraction in right image; (e) disparity map of SGM algorithm based on mutual information; (f) disparity map of proposed algorithm; (g) disparity map of WACS algorithm; (h) rendering disparity map of proposed algorithm; (i) refined disparity map

Download full size

Fig. 6. Experiment results in SuperView-1 satellite stereo pair. (a) Left image; (b) right image; (c) feature point extraction in left image; (d) feature point extraction in right image; (e) disparity map of SGM algorithm based on mutual information; (f) disparity map of proposed algorithm; (g) disparity map of WACS algorithm; (h) rendering disparity map of proposed algorithm; (i) refined disparity map

Download full size

Fig. 7. Accuracy of stereo matching in experimental data set

Download full size

Table 1. Detection results of experimental image feature points
View table
Table 1. Detection results of experimental image feature points
Data set Number of feature points(left image/right image) Number of matching point pairs RMSE
WorldView 601/713 198 0.912
IKONOS 598/655 121 0.896
SuperView-1 523/618 172 0.908

Table 2. Result comparison of different algorithms in experimental images

View table

Table 2. Result comparison of different algorithms in experimental images

Data set	Parameter	SGM^[10]	WACS^[12]	Proposed algorithm
	A₁	0.8723	0.8831	0.9228
WorldView	A₂	0.823	0.825	0.912
	A₃/pixel	5.83	5.04	4.32
	A₁	0.8887	0.8752	0.9587
IKONOS	A₂	0.831	0.834	0.918
	A₃/pixel	3.71	3.72	2.12
	A₁	0.8021	0.8152	0.9621
SuperView-1	A₂	0.816	0.805	0.921
	A₃/pixel	3.82	3.23	2.72

Tools

Get Citation

Copy Citation Text

Yangping Wang, Anna Qin, Qi Hao, Jianwu Dang. Semi-Global Stereo Matching of Remote Sensing Images Combined with Speeded up Robust Features[J]. Acta Optica Sinica, 2020, 40(16): 1628003

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites