Building feature extraction based on natural neighborhood decomposable point feature extraction algorithm

Abstract

Intelligent buildings, emerging as a fusion of modern information technology and architectural structures, aim to provide intelligent, comfortable, and efficient building environments. However, the information processing and decision-making processes within intelligent buildings face challenges in dealing with large-scale, complex data for feature extraction. This research introduces a Natural Neighborhood Decomposable Point algorithm, effectively extracting crucial features within intelligent buildings. This extraction supports the decision-making and control processes for intelligence. Experimental results demonstrate that employing this algorithm for registration reduces registration errors by 0.068 mm and 0.021 mm after the first iteration. This outcome validates the algorithm's efficiency in enhancing registration algorithms, maintaining both accuracy and interpretability while extracting data in intelligent building contexts. This algorithm can effectively analyze both local and global features in point cloud data, contributes to enhancing the energy efficiency, comfort, automation, and intelligence management levels within intelligent buildings. It propels innovation and development in intelligent buildings, improves security, and promotes sustainable development.

翻译

搜索

复制