Remote Sensing Image Change Detection Method Based on Adaptive Boundary Sensing

Fig. 7. Visualization results of each method for images in CLCD dataset. (a) Time 1; (b) time 2; (c) truth; (d) FC-EF; (e) FC-Siam-conc;(f) FC-Siam-diff; (g) STANet; (h) EGRCNN; (i) BIT; (j) MSCANet; (k) ours

Download full size

Fig. 8. Visualization results of each method for images in RSCD dataset. (a) Time 1; (b) time 2; (c) truth; (d) FC-EF; (e) FC-Siam-conc;(f) FC-Siam-diff; (g) STANet; (h) EGRCNN; (i) BIT; (j) MSCANet; (k) ours

Download full size

Fig. 9. Visualization results of ablation experiments. (a) Time 1; (b) time 2; (c) truth; (d) base; (e) base+DE; (f) base+boundary;(g) base+DE+boundary

Download full size

Fig. 10. Visualization results of LEVIRCD dataset. (a) Time 1; (b) time 2; (c) truth; (d) EGRCNN method; (e) ours

Download full size

Table 1. Accuracy comparison of each method on the CLCD dataset

View table

Table 1. Accuracy comparison of each method on the CLCD dataset

Method	Precision /%	Recall /%	F₁ /%	IOU /%
FC-EF	63.96	49.91	56.07	66.46
FC-Siam-conc	59.56	56.39	57.93	67.21
FC-Siam-diff	66.01	45.01	53.52	65.26
STANet	65.13	69.84	67.34	72.71
EGRCNN	59.36	73.27	65.58	71.37
BIT	68.98	55.11	61.27	69.38
MSCANet	71.96	71.46	71.71	75.73
Ours	73.52	72.14	72.82	76.51

Table 2. Accuracy comparison of each method on the RSCD dataset

View table

Table 2. Accuracy comparison of each method on the RSCD dataset

Method	Precision /%	Recall /%	F₁ /%	IOU /%
FC-EF	53.99	29.15	37.85	60.15
FC-Siam-conc	56.18	34.54	42.78	62.13
FC-Siam-diff	58.55	32.75	42.01	61.85
STANet	47.40	56.51	51.62	65.73
EGRCNN	53.26	58.04	55.54	67.74
BIT	61.97	55.88	58.77	69.55
MSCANet	61.37	53.44	57.13	68.71
Ours	63.29	55.19	58.96	69.67

Table 3. Results of ablation experiment using proposed method
View table
Table 3. Results of ablation experiment using proposed method
Method Precision /% Recall /% F₁ /% IOU /%
Base 71.14 68.03 69.56 74.32
Base+DE 69.04 75.18 71.98 75.81
Base+boundary 73.03 71.81 72.42 76.23
Base+DE+boundary 73.52 72.14 72.82 76.51

Table 4. Accuracy comparison of each method on the LEVIRCD dataset

View table

Table 4. Accuracy comparison of each method on the LEVIRCD dataset

Method	Precision /%	Recall /%	F₁ /%	IOU /%
FC-EF	86.91	80.17	83.40	71.53
FC-Siam-conc	91.99	76.77	83.69	71.96
FC-Siam-diff	89.53	83.31	86.31	75.91
STANet	83.81	91.00	87.26	77.39
EGRCNN	87.87	90.80	89.31	89.79
BIT	89.24	89.37	89.30	80.68
Ours	91.55	90.41	90.98	91.25

Tools

Get Citation

Copy Citation Text

Yong Liu, Haitao Guo, Jun Lu, Xiangyun Liu, Lei Ding, Kun Zhu, Donghang Yu. Remote Sensing Image Change Detection Method Based on Adaptive Boundary Sensing[J]. Acta Optica Sinica, 2024, 44(18): 1828001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites