Researching | Fast Image-Stitching Algorithm of Dangerous Objects Under Vehicles

Laser & Optoelectronics Progress, Volume. 61, Issue 8, 0837011(2024)

Fast Image-Stitching Algorithm of Dangerous Objects Under Vehicles

Jianjun Zhang^* and Xin Jin

Author Affiliations

School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, Jiangsu , China

show less

Abstract Get PDF(in Chinese)

Figures&Tables (12)

References (20)

Paper Information

Figures & Tables(12)

Fig. 1. Flow chart of fast stitching algorithm for dangerous images under car

Download full size

Fig. 2. Schematic diagram of FAST feature point detection algorithm

Download full size

Fig. 3. Flow chart of FAST feature point extraction

Download full size

Fig. 4. Three sets of images to be spliced. (a) Image group 1; (b) image group 2; (c) image group 3

Download full size

Fig. 5. Image feature point matching renderings of four methods. (a) Image group 1; (b) image group 2

Download full size

Fig. 6. Image fusion effect comparison. (a) Image group 1; (b) image group 2

Download full size

Fig. 7. Two images mosaic comparison experiment. (a) Image group 1; (b) image group 2; (c) image group 3

Download full size

Fig. 8. Multiple image stitching renderings

Download full size

Table 1. Image feature point matching indexes of four algorithms in image group 1 and 2

Table 1. Image feature point matching indexes of four algorithms in image group 1 and 2

	Image group 1				Image group 2
Index	SIFT	SURF	ORB	Proposed algorithm	SIFT	SURF	ORB	Proposed algorithm
Fpdt /s	0.491	0.672	0.361	0.342	0.376	0.501	0.354	0.318
Coamc	1425	2007	501	378	299	515	410	367
Cormc	1097	1709	435	349	203	354	253	289
Fpmt /s	1.194	1.290	0.237	0.223	0.264	0.634	0.089	0.076
Ma /%	76.98	85.15	86.82	92.33	67.89	68.73	61.70	78.74

Table 2. Experimental comparison of purification of feature points in image groups 1 and 2

Table 2. Experimental comparison of purification of feature points in image groups 1 and 2

	Image group 1				Image group 2
Index	ORB+ RANSAC	ORB+ PROSAC	FAST+BIRSK+ RANSAC	Proposed algorithm	ORB+ RANSAC	ORB+ PROSAC	FAST+BRISK+ RANSAC	Proposed algorithm
Cormc	501	501	378	378	410	410	367	367
Amc	104	119	82	105	43	56	48	68
Fpur /%	20.76	23.75	21.69	27.77	10.48	13.65	13.07	18.52
Fprt /s	0.235	0.213	0.219	0.138	0.185	0.157	0.158	0.102

Table 3. Ablation experimental data

Table 3. Ablation experimental data

Algorithm	Fpdt /s	Fpmt /s	Ma /%	Fpur /%	St /s
①ORB	0.231	0.133	73.33	18.91	0.362
②FAST+BRISK	0.189	0.085	86.54
③FAST+BRISK+RANSAC	0.190	0.087	86.11	20.15
④FAST+BRISK+PROSAC	0.189	0.086	87.82	25.58
⑤FAST+BRISK+PROSAC+ Weighted average fusion	0.189	0.086	87.82	25.58	0.328
⑥Proposed algorithm	0.189	0.086	87.82	25.58	0.297

Table 4. Two image stitching experimental data

Table 4. Two image stitching experimental data

Image grouping	Algorithm	Is there any ghosting	Visible stitching seams	E_RMSE	Stitching time /s
Image group 1	SIFT	No	Yes	2.06	1.715
	SURF	Yes	Yes	2.08	1.719
	ORB	No	Yes	1.91	0.759
	Proposed algorithm	No	No	1.19	0.599
Image group 2	SIFT	No	Yes	2.42	0.765
	SURF	No	Yes	1.95	0.751
	ORB	No	Yes	2.19	0.486
	Proposed algorithm	No	No	1.47	0.383
Image group 3	SIFT	Yes	Yes	1.87	1.248
	SURF	No	Yes	2.08	1.381
	ORB	Yes	No	1.75	0.362
	Proposed algorithm	No	No	1.41	0.297

Tools

Get Citation

Copy Citation Text

Jianjun Zhang, Xin Jin. Fast Image-Stitching Algorithm of Dangerous Objects Under Vehicles[J]. Laser & Optoelectronics Progress, 2024, 61(8): 0837011

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Digital Image Processing

Received: Jul. 3, 2023

Accepted: Aug. 8, 2023

Published Online: Apr. 2, 2024

The Author Email: Zhang Jianjun (jjzhuang@nuist.edu.cn)

DOI:10.3788/LOP231652

Topics

laser devices and laser physics

Lasers and Laser Optics

laser manufacturing

Instrumentation, Measurement and Metrology