Abstract
In this article, a remote sensing image change detection method based on depthwise separable convolution with U-Net is proposed, which omits the tedious steps of generating and analyzing the difference map in the traditional remote sensing image change detection method. First, two images having c-channel each can be specifically stacked into a 2c-channel image, and the change detection can be converted to an image segmentation problem, an improved full convolution network (FCN) called U-Net is exploited to directly separate the changing regions. Because the capability of the deep convolution network is proportional to the depth of the network and a deeper convolution network means the increase of the training parameters, we then replace the original convolution in FCN by the depthwise separable convolution, making the entire network lighter, while the model performs slightly better than the traditional convolution operation. Besides that, another innovation in our proposed method is to use a preference control loss function to meet the different needs of precision and recall rate. Experimental results validate the effectiveness and robustness of the proposed method.
Highlights
I MAGE change detection is to detect the change of the two images taken at different times in the same place
Gong et al [4] proposed an unsupervised method based on deep neural networks (DNN), which can obtain final change detection map directly from the two original images
Since the binary image segmentation can be regarded as a pixel-level classification task, the common loss function of the binary image classification problem can be applied during network training, which is called binary cross entropy (BCE)
Summary
I MAGE change detection is to detect the change of the two images taken at different times in the same place. The parcel-based change detection method uses the object as the analysis unit [3], which requires relatively low registration accuracy It can directly obtain the change target and facilitate subsequent processing. Gong et al [4] proposed an unsupervised method based on deep neural networks (DNN), which can obtain final change detection map directly from the two original images. Mou et al [12] proposed a novel network architecture, which is trained to learn a joint spectral–spatial–temporal feature representation in a unified framework for change detection of multispectral images. For this purpose, they combined CNN and RNN into an end-to-end network framework.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
