Building Outline Extraction Directly Using the U2-Net Semantic Segmentation Model from High-Resolution Aerial Images and a Comparison Study

Abstract

Deep learning techniques have greatly improved the efficiency and accuracy of building extraction using remote sensing images. However, high-quality building outline extraction results that can be applied to the field of surveying and mapping remain a significant challenge. In practice, most building extraction tasks are manually executed. Therefore, an automated procedure of a building outline with a precise position is required. In this study, we directly used the U2-net semantic segmentation model to extract the building outline. The extraction results showed that the U2-net model can provide the building outline with better accuracy and a more precise position than other models based on comparisons with semantic segmentation models (Segnet, U-Net, and FCN) and edge detection models (RCF, HED, and DexiNed) applied for two datasets (Nanjing and Wuhan University (WHU)). We also modified the binary cross-entropy loss function in the U2-net model into a multiclass cross-entropy loss function to directly generate the binary map with the building outline and background. We achieved a further refined outline of the building, thus showing that with the modified U2-net model, it is not necessary to use non-maximum suppression as a post-processing step, as in the other edge detection models, to refine the edge map. Moreover, the modified model is less affected by the sample imbalance problem. Finally, we created an image-to-image program to further validate the modified U2-net semantic segmentation model for building outline extraction.

Enhanced Semantic Information Transfer of Multi-Domain Samples: An Adversarial Edge Detection Method Using Few High-Resolution Remote Sensing Images

“…The development of edge detection methods based on deep learning has brought more solutions to the information extraction of high-resolution remote sensing images. Wei et al [ 29 ] used the U2-net [ 30 ] semantic segmentation model to detect building edges and replaced the original loss function with a multi-class cross-entropy loss function to directly generate a binary map with edges and backgrounds. Xia et al [ 31 ] proposed a building edge detection method that uses Faster R-CNN [ 32 ] to detect the bounding box of the building and uses the bounding box to assist in the repair of the broken line to completely extract the building outline.…”

Automatic Extraction of the Calving Front of Pine Island Glacier Based on Neural Network

“…The model incorporates a two-layer nested 'U' shaped structure and introduces the concept of Residual U-blocks (RSU). Comparative experiments involving networks of the same architecture have demonstrated that U 2 -Net not only enhances network depth but also maintains lower computational costs while effectively capturing image information at different scales [35,39,40]. In this paper, U 2 -Net is utilized to obtain classification results for the ice shelf and ocean regions.…”

Highly Accurate Dichotomous Image Segmentation

“…Image segmentation, as one of the most fundamental techniques in computer vision, plays a vital role in enabling the machines to perceive and understand the real world. Compared with image classification [17,47,84] and object detection [33,34,80], it can provide more geometrically accurate descriptions of the targets used in a wide range of applications, such as image editing [36], 3D reconstruction [57], augmented reality (AR) [76], satellite image anal-ysis [100], medical image processing [81], robot manipulation [8], etc. We can categorize the above applications as "light" (e.g., image editing and image analysis) and "heavy" (e.g., manufacturing and surgical robots), based on their immediate affects on real-world objects.…”