Name
Title | ||
|---|---|---|
Change Detection in Synthetic Aperture Radar Images based on Convolutional Block Attention Module |
Abstract | ||
|---|---|---|
Change detection with SAR images encounters more difficulties than optical ones due to the presence of the speckle noise. Recently, inspired by the human visual systems, self-attention mechanism is incorporated in large-scale classification tasks to improve the classification performance. Inspired by this, this paper proposed a novel change detection method from SAR images based on self-attention mechanism. First, training samples with pseudo-labels are generated by the log-ratio operator and hierarchical FCM algorithm. Image patches around these pixel are generated, and these patches are fed into a network with the convolutional block attention module (CBAM). After training, all the pixels from multitemporal SAR images are fed into the learned network, and then the final change map can be obtained. Experimental results on two real SAR datasets demonstrate the superiority of the proposed method over two closely related methods. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/Multi-Temp.2019.8866962 | 2019 10th International Workshop on the Analysis of Multitemporal Remote Sensing Images (MultiTemp) |
Keywords | Field | DocType |
Synthetic aperture radar,change detection,clustering method,convolutional block attention module | Computer vision,Change detection,Speckle pattern,Synthetic aperture radar,Computer science,Pixel,Artificial intelligence,Operator (computer programming),Speckle noise | Conference |
ISBN | Citations | PageRank |
978-1-7281-4616-4 | 0 | 0.34 |
References | Authors | |
8 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Dong Wang | 1 | 1351 | 186.07 |
| feng gao | 2 | 53 | 17.81 |
| Junyu Dong | 3 | 393 | 77.68 |
| Shengke Wang | 4 | 1 | 1.70 |