PCBA-Net: Pyramidal Convolutional Block Attention Network for Synthetic Aperture Radar Image Change Detection

Synthetic aperture radar (SAR) imagery change detection (CD) is still a crucial and challenging task. Recently, with the boom of deep learning technologies, many methods have been presented for SAR CD, they achieve superior performance to traditional methods. However, most available convolutional neural networks (CNN) approaches use diminutive single convolution kernel, which has small receptive field cannot make full context information some useful detail images. In order address above drawback, pyramidal block attention network (PCBA-Net) proposed image CD in this study. The PCBA-Net consists (PyConv) module (CBAM). PyConv can not only extend input capture enough information, but also handles incremental kernel sizes parallel obtain multi-scale detailed information. Additionally, CBAM introduced emphasize To verify our method, six actual datasets are used experiments. results real reveal that approach outperforms several state-of-the-art