User-Assisted Procedural Facade Encoding

Huilong Zhuo, Purdue University

Document Type Other

Abstract

This study proposes and develops a new facade encoding tool. The unique feature of this tool is the engagement of user interaction in the encoding process. The user-assisted encoding tool is developed in a way that gives the user the freedom to observe and decide how to encode a facade as a set of procedural rules. The study has the goal to represent a building facade by its grammar with user interactivity and thus compress the facade model, so that computer resources can be saved as a result of the encoding. The tool is developed using OpenGL and Qt SDK libraries. It enables the user to decide the splitting rules to encode the facade. It can also reconstruct a building facade based on the resulting terminal facade, as well as calculate the compression factor, which is the measurement of the efficiency of the ompression process, and finally the decompression overhead time and rendering time after compression. Seven cases of building facades are put into test after the tool is developed. The researcher of this study acts as the user to decide the splitting rule for each facade. Two approaches-vertical and horizontal split-are adopted. The results show that, through such a user-assisted approach, the average compression factor is lower than 1/3, thus can save over 2/3 of the space needed for storing the image on average. Also, the decompression and rendering time together are shorter than the loading and rendering time for the uncompressed facade. Therefore the advantages in both the compression factor and time indicate the strength of such a user-assisted encoding tool.