STCD-EffV2T Unet: Semi Transfer Learning EfficientNetV2 T-Unet Network for Urban/Land Cover Change Detection Using Sentinel-2 Satellite Images

There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

Abstract

Change detection in urban areas can be helpful for urban resource management and smart city planning. The effects of human activities on the environment and ground have gained momentum over the past decades, causing remote sensing data sources analysis (such as satellite images) to become an option for swift change detection in the environment and urban areas. We proposed a semi-transfer learning method of EfficientNetV2 T-Unet (EffV2 T-Unet) that combines the effectiveness of composite scaled EfficientNetV2 T as the first path or encoder for feature extraction and convolutional layers of Unet as the second path or decoder for reconstructing the binary change map. In the encoder path, we use EfficientNetV2 T, which was trained by the ImageNet dataset. In this research, we employ two datasets to evaluate the performance of our proposed method for binary change detection. The first dataset is Sentinel-2 satellite images which were captured in 2017 and 2021 in urban areas of northern Iran. The second one is the Onera Satellite Change Detection dataset (OSCD). The performance of the proposed method is compared with YoloX-Unet families, ResNest-Unet families, and other well-known methods. The results demonstrated our proposed method’s effectiveness compared to other methods. The final change map reached an overall accuracy of 97.66%.

Related collections

Most cited references 55

Record: found
Abstract: found
Article: not found

DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs

Kevin Murphy, Iasonas Kokkinos, George Papandreou … (2018)

In this work we address the task of semantic image segmentation with Deep Learning and make three main contributions that are experimentally shown to have substantial practical merit. First, we highlight convolution with upsampled filters, or 'atrous convolution', as a powerful tool in dense prediction tasks. Atrous convolution allows us to explicitly control the resolution at which feature responses are computed within Deep Convolutional Neural Networks. It also allows us to effectively enlarge the field of view of filters to incorporate larger context without increasing the number of parameters or the amount of computation. Second, we propose atrous spatial pyramid pooling (ASPP) to robustly segment objects at multiple scales. ASPP probes an incoming convolutional feature layer with filters at multiple sampling rates and effective fields-of-views, thus capturing objects as well as image context at multiple scales. Third, we improve the localization of object boundaries by combining methods from DCNNs and probabilistic graphical models. The commonly deployed combination of max-pooling and downsampling in DCNNs achieves invariance but has a toll on localization accuracy. We overcome this by combining the responses at the final DCNN layer with a fully connected Conditional Random Field (CRF), which is shown both qualitatively and quantitatively to improve localization performance. Our proposed "DeepLab" system sets the new state-of-art at the PASCAL VOC-2012 semantic image segmentation task, reaching 79.7 percent mIOU in the test set, and advances the results on three other datasets: PASCAL-Context, PASCAL-Person-Part, and Cityscapes. All of our code is made publicly available online.