High-Resolution Remote Sensing Scene Classification Based on Salient Features and DCNN

Huanhuan Lü; Tao Liu; Hui Zhang; Guofeng Peng; Juntong Zhang

doi:10.3788/LOP202158.2001002

Laser & Optoelectronics Progress, Volume. 58, Issue 20, 2001002(2021)

High-Resolution Remote Sensing Scene Classification Based on Salient Features and DCNN

Huanhuan Lü^*, Tao Liu^**, Hui Zhang, Guofeng Peng, and Juntong Zhang

Author Affiliations

College of Software, Liaoning Technical University, Huludao, Liaoning 125105, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(16)

Fig. 1. Flow chart of our method

Download full size

Fig. 2. Segmentation results with different number of superpixels

Download full size

Fig. 3. Extraction result of saliency map. (a) Original image; (b) K-means clustering; (c) superpixel segmentation; (d) fusion result

Download full size

Fig. 4. Extraction result of the ROI. (a) Saliency map;(b) gray enhancement map; (c) binarization map; (d) ROI

Download full size

Fig. 5. Structure of the DCNN

Download full size

Fig. 6. Expanded results of the data. (a) Original image; (b) horizontal flip; (c) vertical flip; (d) brightness adjustment

Download full size

Fig. 7. Loss function and classification accuracy of different models. (a) UC-Merced data set; (b) WHU-RS data set

Download full size

Fig. 8. Images in different data sets. (a) UC-Merced data set; (b) WHU-RS data set

Download full size

Fig. 9. Experimental results of the UC-Merced data set. (a) Original image; (b) saliency map; (c) gray enhancement map; (d) binarization map; (e) ROI

Download full size

Fig. 10. Confusion matrix of our method on the UC-Merced data set

Download full size

Fig. 11. Experimental results in the WHU-RS data set. (a) Original image; (b) saliency map; (c) gray enhancement map; (d) binarization map; (e) ROI

Download full size

Fig. 12. Confusion matrix of our method on the WHU-RS data set

Download full size

Table 1. Classification accuracy of the UC-Merced data set unit: %

View table

Table 1. Classification accuracy of the UC-Merced data set unit: %

No.	Scene	Accuracy	No.	Scene	Accuracy	No.	Scene	Accuracy
1	agricultural	96	8	forest	95	15	overpass	100
2	airplane	93	9	freeway	100	16	parking lot	100
3	baseball diamond	100	10	golf course	100	17	river	100
4	beach	100	11	harbor	100	18	runway	96
5	buildings	84	12	intersection	100	19	sparse residential	100
6	chaparral	100	13	medium residential	84	20	storage tanks	88
7	dense residential	92	14	mobile home park	100	21	tennis court	90
Average accuracy					96.10

Table 2. Classification accuracies of different methods on the UC-Merced data set
View table
Table 2. Classification accuracies of different methods on the UC-Merced data set
Method Accuracy /% Time /s
Saliency + Multi-CNN 92.86 2.27
MS-DCNN 91.34 5.90
JMCNN 88.30 0.81
Our method 96.10 1.95

Table 3. Classification accuracy of the WHU-RS data set unit: %

View table

Table 3. Classification accuracy of the WHU-RS data set unit: %

No.	Scene	Accuracy	No.	Scene	Accuracy	No.	Scene	Accuracy
1	airport	100	8	football field	100	15	pond	100
2	beach	100	9	forest	92	16	port	100
3	bridge	100	10	industrial	93	17	rail way station	93
4	commercial	91	11	meadow	86	18	residential	91
5	desert	100	12	mountain	100	19	river	92
6	farmland	91	13	park	100
7	viaduct	92	14	parking	100
Average accuracy						95.84

Table 4. Classification accuracies of different methods on the WHU-RS data set
View table
Table 4. Classification accuracies of different methods on the WHU-RS data set
Method Accuracy /% Time /s
Saliency + Multi-CNN 91.80 3.13
MS-DCNN 90.05 7.33
JMCNN 87.63 0.98
Our method 95.84 2.32

Tools

Get Citation

Copy Citation Text

Huanhuan Lü, Tao Liu, Hui Zhang, Guofeng Peng, Juntong Zhang. High-Resolution Remote Sensing Scene Classification Based on Salient Features and DCNN[J]. Laser & Optoelectronics Progress, 2021, 58(20): 2001002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites