Super-Resolution-Based Snake Model—An Unsupervised Method for Large-Scale Building Extraction Using Airborne LiDAR Data and Optical Image

Abstract

Automatic extraction of buildings in urban and residential scenes has become a subject of growing interest in the domain of photogrammetry and remote sensing, particularly since the mid-1990s. Active contour model, colloquially known as snake model, has been studied to extract buildings from aerial and satellite imagery. However, this task is still very challenging due to the complexity of building size, shape, and its surrounding environment. This complexity leads to a major obstacle for carrying out a reliable large-scale building extraction, since the involved prior information and assumptions on building such as shape, size, and color cannot be generalized over large areas. This paper presents an efficient snake model to overcome such a challenge, called Super-Resolution-based Snake Model (SRSM). The SRSM operates on high-resolution Light Detection and Ranging (LiDAR)-based elevation images—called z-images—generated by a super-resolution process applied to LiDAR data. The involved balloon force model is also improved to shrink or inflate adaptively, instead of inflating continuously. This method is applicable for a large scale such as city scale and even larger, while having a high level of automation and not requiring any prior knowledge nor training data from the urban scenes (hence unsupervised). It achieves high overall accuracy when tested on various datasets. For instance, the proposed SRSM yields an average area-based Quality of 86.57% and object-based Quality of 81.60% on the ISPRS Vaihingen benchmark datasets. Compared to other methods using this benchmark dataset, this level of accuracy is highly desirable even for a supervised method. Similarly desirable outcomes are obtained when carrying out the proposed SRSM on the whole City of Quebec (total area of 656 km2), yielding an area-based Quality of 62.37% and an object-based Quality of 63.21%.

Multi-Building Extraction and Alignment for As-Built Point Clouds: A Case Study With Crane Cameras

“…Building extraction techniques have largely been developed for distant aerial imagery for urban planning and monitoring applications (Tomljenovic et al, 2015;Boonpook et al, 2020). Building extraction can either be data-driven (Nguyen et al, 2020) or model-driven (Zheng and Weng, 2015). Data-driven methods are the natural choice when a suitable model is not present.…”

Building Extraction from Airborne Multi-Spectral LiDAR Point Clouds Based on Graph Geometric Moments Convolutional Neural Networks

“…Nguyen et al [29] presented a super-resolution-based snake model (SRSM) to extract buildings using LiDAR data and optical image. First, they also preliminary extracted the candidate building points by using the DTM and NDVI derived from LiDAR data and optical image, respectively.…”

Snake-Based Model for Automatic Roof Boundary Extraction in the Object Space Integrating a High-Resolution Aerial Images Stereo Pair and 3D Roof Models

“…Several building extraction methods have been reported due to the emergence of different remote sensing technology over the last few decades. For example, aerial images have been used in many works to extract buildings [4][5][6][7], but they face many problems due to occlusions, shadows, and disadvantageous image perspectives [8]. On the other hand, other works [9][10][11][12] have proposed the use of light detection and ranging (LiDAR) data to exploit the 3D information for building extraction.…”

“…The main disadvantage of MorphoGACs is the need for strong edges, making it necessary to use gradient images, for example. Nguyen et al [8] presented the super-resolution-based Snake model (SRSM) for large-scale building extraction. The SRSM operates on LiDAR-based elevation images moving the Snake curve adaptively based on a pre-provided building mask.…”