SCD-Net: A Semantic–Contour–Disparity Triadic Collaborative Network for 3-D Building Reconstruction From Stereo Satellite Imagery

Abstract

For the task of 3-D building reconstruction from stereo satellite imagery, existing methods often treat semantic segmentation, contour regularization, and disparity estimation as independent or loosely coupled tasks, failing to fully exploit their geometric correlations. To address this, this letter proposes the semantic–contour–disparity triadic collaborative network (SCD-Net). This network establishes a deeply coupled computational framework that extracts multiscale features through a shared encoder, subsequently routing them to dedicated branches. Specifically, we design a multiscale edge aggregation (MEA) module that significantly enhances the refined representation of building edges, and also propose a SCD Upsampling (SCDU) module, which leverages dual geometric priors from semantic categories and building contours to guide disparity optimization, effectively resolving the issue of blurred building boundaries in the disparity map. Furthermore, a novel building contour loss function is introduced, strengthening multitask constraints by quantifying the geometric consistency between disparity edges and semantic contours. Experiments on the US3D and WHU datasets demonstrate that the SCD-Net achieves state-of-the-art (SOTA) performance on key metrics for stereo matching and building semantic segmentation.