Automated building facade quality assessment using 3D laser scanning and improved k-d tree–accelerated point-cloud analysis

Facade inspection during construction is critical for quality control and compliance with design and regulatory requirements, yet conventional methods are labor-intensive, time-consuming, and prone to human error. This study proposes an automated framework for assessing facade quality during the construction phase, based on 3D laser scanning and an enhanced k-d tree–accelerated point-cloud analysis approach. The method enhances the computational efficiency of boundary feature point extraction by constructing a k-d tree, enabling rapid feature recognition of the facade point cloud. The extracted structural features are then modeled using the least squares method to detect facade deformation. To replicate manual inspection logic, a virtual straightedge sliding mechanism is introduced, enabling quantitative evaluation of facade flatness and verticality through localized fitting and deviation analysis. Validation using real construction projects demonstrates that the proposed framework achieves higher efficiency, accuracy, and robustness, delivering high-resolution, traceable, and engineering-ready results for digital construction quality management. Copyright © 2026. Published by Elsevier Ltd.