Classification of Forest Types using UAV Remote Sensing Images Based on Improved Ant Colony Algorithm

Guiling Zhao; Pengnian Li; Quanrong Guo; Maolin Tan

doi:10.3788/LOP202259.1628002

Laser & Optoelectronics Progress, Volume. 59, Issue 16, 1628002(2022)

Classification of Forest Types using UAV Remote Sensing Images Based on Improved Ant Colony Algorithm

Guiling Zhao, Pengnian Li^*, Quanrong Guo, and Maolin Tan

Author Affiliations

School of Geomatics, Liaoning Technical University, Fuxin 123000, Liaoning , China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(11)

Fig. 1. Classification effect under test set. (a) Blood; (b) Vehicle; (c) Statlog; (d) Glass; (e) Haberman

Download full size

Fig. 2. Geographical location of the study area

Download full size

Fig. 3. Gabor filter banks and processing results of different ground features. (a) Filter bank; (b) needle-broad-leaved mixed forest; (c) coniferous forest; (d) broad-leaved forest; (e) bare land; (f) water

Download full size

Fig. 4. Diagram of owner principal component analysis. (a) Needle-broad-leaved mixed forest; (b) coniferous forest; (c) broad-leaved forest; (d) bare land; (e) water bodies

Download full size

Fig. 5. 5 types of ground feature samples. (a) Needle-broad-leaved mixed forest; (b) coniferous forest; (c) broad-leaved forest; (d) bare land; (e) water

Download full size

Fig. 6. Confusion matrix of recognition results without texture features

Download full size

Fig. 7. Confusion matrix of recognition results under GLMC feature

Download full size

Fig. 8. Confusion matrix of recognition results under Gabor feature

Download full size

Table 1. Public dataset properties
View table
Table 1. Public dataset properties
Group Category Number of categories Number of instances Data length
1 Blood 2 4 748
2 Vehicle 2 13 270
3 Statlog 3 4 150
4 Glass 6 9 214
5 Haberman 2 3 306

Table 2. Optimal classification results of five data sets

View table

Table 2. Optimal classification results of five data sets

Category	Algorithm	$c$	$γ$	Precision /%	Running time /s
Blood	ABC-SVM	4.9257	0.2862	77.15	4.13
	GA-SVM	4.5329	0.8725	77.42	5.26
	ACO-SVM	8.3513	0.9097	77.78	10.07
Vehicle	ABC-SVM	3.3783	0.7945	83.67	5.30
	GA-SVM	6.1811	0.6184	84.68	3.86
	ACO-SVM	9.0413	0.9912	86.69	7.82
Statlog	ABC-SVM	3.4880	0.8489	92.02	2.50
	GA-SVM	7.8337	0.4546	90.08	2.35
	ACO-SVM	9.3614	0.8992	96.00	4.92
Glass	ABC-SVM	9.8256	0.5255	98.83	2.17
	GA-SVM	6.0869	0.5578	98.24	1.75
	ACO-SVM	9.3535	0.6589	99.41	4.98
Haberman	ABC-SVM	4.2138	0.9758	77.7	2.39
	GA-SVM	7.4841	0.6786	75.8	3.25
	ACO-SVM	8.5761	0.9503	78.3	6.08

Table 3. Recognition accuracy of different classification models
View table
Table 3. Recognition accuracy of different classification models
Algorithm Number of samples Number of correctly classified samples Number of incorrectly classified samples Overall accuracy /%
ACO-SVM 200 162 38 81
GA-SVM 200 158 42 79
ABC-SVM 200 157 43 78.5
SVM 200 149 51 74.5

Tools

Get Citation

Copy Citation Text

Guiling Zhao, Pengnian Li, Quanrong Guo, Maolin Tan. Classification of Forest Types using UAV Remote Sensing Images Based on Improved Ant Colony Algorithm[J]. Laser & Optoelectronics Progress, 2022, 59(16): 1628002

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites