High resolution building facade dataset for facade elements parsingMendeley Data
This paper introduces a high-resolution building facade dataset—hznu_facade, designed to provide rich data support for urban modeling, digital twin city construction, and facade analysis tasks. The dataset comprises facade images of 624 buildings from Hangzhou, China, covering diverse architectural...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-10-01
|
| Series: | Data in Brief |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2352340925006742 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This paper introduces a high-resolution building facade dataset—hznu_facade, designed to provide rich data support for urban modeling, digital twin city construction, and facade analysis tasks. The dataset comprises facade images of 624 buildings from Hangzhou, China, covering diverse architectural styles from both commercial and residential areas. Additionally, it offers twin versions of the facade images with perspective correction. To ensure data diversity, the dataset includes images captured under various lighting conditions, including challenging environments such as direct sunlight and backlighting. Furthermore, the dataset places particular emphasis on the annotation of window elements, offering high-density window information that significantly improves the capabilities for architectural element recognition and analysis. During the annotation process, this paper leverages existing segmentation models to assist in the labelling of architectural elements, covering six primary categories: buildings, cars, trees, windows, skies, and doors. The dataset annotates a total of 43,277 windows, far exceeding existing related datasets, highlighting its strong potential for building analysis, architectural element recognition, and virtual reality (VR) applications. This dataset provides an valuable benchmark for facade parsing algorithms, contributing to the advancement of research and development in this field. |
|---|---|
| ISSN: | 2352-3409 |