Innovative Applications of Intelligent Assistive Systems in Interior Design

Abstract

With the development of smart homes, virtual reality, and other technologies, people’s demand for digital modeling and visualization of interior space is increasing. How to quickly and accurately estimate the layout of the interior space has become a matter of great concern. This paper develops a 3D virtual visual interior design platform structure based on this. By studying the indoor multi-sensor fusion localization system, the localization module in the navigation process was designed based on this fusion result. The visual localization algorithm, WiFi localization, PDR localization, and AMCL fusion localization algorithms are introduced, respectively. The extended Kalman filter algorithm is used to fuse the algorithm with the highest state value for localization, which is used for scene recognition according to the algorithm localization effect. To achieve the 3D virtual visual impact, a 3D virtual visual interior design platform is created based on 3D virtual vision. Experiments are conducted on the localization system in the platform, and the error of all the localization points of the localization method combining vision and AMCL is no more than 1.1m, with the maximum error of 1.071m and the average error of 0.481m, which is the lowest error among all the algorithms. The localization sensor constructed by the localization algorithm using vision and AMCL with extended Kalman filtering has an average error of 10.33cm, which allows for good scene recognition. The data analyzes designers’ satisfaction with the use of the platform, and the independent variable in the regression equation that contributes most significantly to total satisfaction is the platform’s time-saving, with a coefficient of 0.324. The platform’s focus should be on enhancing time efficiency.

Keywords

• 3d virtual vision
• localization sensor
• visual localization algorithm
• amcl algorithm
• interior design
• 97p20