Road Extraction Method from Remote Sensing Images with Feature Consistency Perception

Xuyang Zhao; Feng Luo; Hui Yang; Biao Wang; Guangyao Ren; Yongchuang Wu

doi:10.3788/LOP232635

Laser & Optoelectronics Progress, Volume. 61, Issue 18, 1828002(2024)

Road Extraction Method from Remote Sensing Images with Feature Consistency Perception

Xuyang Zhao¹, Feng Luo², Hui Yang^3、*, Biao Wang¹, Guangyao Ren⁴, and Yongchuang Wu⁵

Author Affiliations

¹School of Resources and Environmental Engineering, Anhui University, Hefei 230601, Anhui, China

²Second Highway Consultants Company Ltd., Wuhan 430056, Hubei, China

³Institutes of Physical Science and Information Technology, Anhui University, Hefei 230601, Anhui, China

⁴Second Surveying and Mapping Institute of Anhui Province, Hefei 230601, Anhui, China

⁵School of Artificial Intelligence, Anhui University, Hefei 230601, Anhui, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(13)

Fig. 1. FECP-Net model structure

Download full size

Fig. 2. DenseASPP model structure

Download full size

Fig. 3. Examples of the CHT dataset

Download full size

Fig. 4. Examples of the Massachusetts dataset

Download full size

Fig. 5. Examples of the DeepGlobal dataset

Download full size

Fig. 6. Comparative experimental results on the CHT dataset, where the wireframe reflects the differences between different methods

Download full size

Fig. 7. Comparative experimental results on the Massachusetts dataset, where the wireframe reflects the differences between different methods

Download full size

Fig. 8. Comparative experimental results on the DeepGlobal dataset, where the wireframe reflects the differences between different methods

Download full size

Fig. 9. Comparative experimental results on the CHT dataset in ablation experiment

Download full size

Table 1. Accuracy evaluation for the comparative experiments in CHT
View table
Table 1. Accuracy evaluation for the comparative experiments in CHT
Model IOU Recall F1
U-Net 67.96 80.65 77.07
D-LinkNet 71.47 83.03 76.63
DGRN 71.70 83.21 80.09
Proposed model 82.78 90.41 85.19

Table 2. Accuracy evaluation for the comparative experiments in Massachusetts
View table
Table 2. Accuracy evaluation for the comparative experiments in Massachusetts
Model IOU Recall F1
U-Net 48.13 68.13 64.99
D-LinkNet 54.25 74.66 70.35
DGRN 57.16 75.99 71.85
Proposed model 57.61 80.59 73.11

Table 3. Accuracy evaluation for the comparative experiments on the DeepGlobal dataset
View table
Table 3. Accuracy evaluation for the comparative experiments on the DeepGlobal dataset
Model IOU Recall F1
U-Net 50.27 61.28 65.36
D-LinkNet 66.17 73.38 79.43
DGRN 67.25 75.01 79.64
Proposed model 72.55 80.18 83.56

Table 4. Accuracy evaluation of ablation experiments on the CHT dataset
View table
Table 4. Accuracy evaluation of ablation experiments on the CHT dataset
Model IOU Recall F1
FECP-Net 85.12 92.36 91.32
CRE-Net 79.28 85.76 84.64

Tools

Get Citation

Copy Citation Text

Xuyang Zhao, Feng Luo, Hui Yang, Biao Wang, Guangyao Ren, Yongchuang Wu. Road Extraction Method from Remote Sensing Images with Feature Consistency Perception[J]. Laser & Optoelectronics Progress, 2024, 61(18): 1828002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites