Feature Point Matching Algorithm for Fusion of Color Invariants and SURB Detection

Shu Chen; Tian Yang; Shunyuan Sun

doi:10.3788/LOP55.051007

Journals >Laser & Optoelectronics Progress >Volume 55 >Issue 5 >Page 051007 > Article

Laser & Optoelectronics Progress
Vol. 55, Issue 5, 051007 (2018)

Feature Point Matching Algorithm for Fusion of Color Invariants and SURB Detection

Shu Chen, Tian Yang^*;, and Shunyuan Sun

Author Affiliations

School of Internet of Things Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China

show less

DOI: 10.3788/LOP55.051007 Cite this Article Set citation alerts

Shu Chen, Tian Yang, Shunyuan Sun. Feature Point Matching Algorithm for Fusion of Color Invariants and SURB Detection[J]. Laser & Optoelectronics Progress, 2018, 55(5): 051007 Copy Citation Text

show less

Fig. 1. FAST corner detection

Download full size

Matching effects of ORB algorithm under different illuminations and scales. (a)(b) Input images under different illuminations; (c) matching effect under illumination changes; (d)(e) input images under different scales; (f) matching effect under scale changes

Fig. 2. Matching effects of ORB algorithm under different illuminations and scales. (a)(b) Input images under different illuminations; (c) matching effect under illumination changes; (d)(e) input images under different scales; (f) matching effect under scale changes

Download full size

Fig. 3. Process of proposed matching algorithm

Download full size

Fig. 4. Image matching effect under different illuminations. (a) Input image; (b) test image 1; (c) test image 2; (d) test image 3; (e) match image 1; (f) match image 2; (g) match image 3

Download full size

Fig. 5. Image matching effect under different angles. (a) Input image; (b) test image 1; (c) test image 2; (d) test image 3; (e) match image 1; (f) match image 2; (g) match image 3

Download full size

Fig. 6. Image matching effect under different scales. (a) ORB algorithm; (b) algorithm in Ref. [13]; (c) proposed algorithm

Download full size

Data	ORB+RANSAC		ORB+optimizedRANSAC		CORB		Algorithm inRef. [13]		Proposed method
Data	T /s	R /%	T /s	R /%	T /s	R /%	T /s	R /%	T /s	R /%
Data 1	0.41	11.42	0.32	5.56	0.46	8.56	10.52	4.21	0.45	4.15
Data 2	0.43	21.58	0.31	11.53	0.47	11.73	10.49	11.38	0.44	8.65
Data 3	0.42	58.40	0.34	22.86	0.47	18.86	10.53	14.07	0.45	10.38

Table 1. Comparison of mis-matching rates of each algorithm under different illuminations

Data	ORB+RANSAC		ORB+optimizedRANSAC		CORB		Algorithm in Ref. [13]		Proposed method
Data	T /s	R /%	T /s	R /%	T /s	R /%	T /s	R /%	T /s	R /%
Data 1	0.41	11.52	0.28	3.16	0.46	8.36	10.49	6.58	0.44	2.50
Data 2	0.32	21.35	0.19	18.74	0.35	4.95	9.46	4.21	0.34	3.48
Data 3	0.34	58.79	0.19	10.86	0.39	8.70	9.68	5.68	0.36	5.31

Table 2. Comparison of mis-matching rates of each algorithm under different rotating angles

Shu Chen, Tian Yang, Shunyuan Sun. Feature Point Matching Algorithm for Fusion of Color Invariants and SURB Detection[J]. Laser & Optoelectronics Progress, 2018, 55(5): 051007

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information