TOWARDS AN EFFICIENT CITY INVENTORY MANAGEMENT SYSTEM FOR URBAN AUTHORITIES IN DEVELOPING COUNTRIES - THE CASE OF 3D CHANGE DETECTION

As cities and urban areas grow and become more complex, the task of managing, updating and displaying three-dimensional urban information can be challenging for city and urban authorities. The emergence of 3D city models has enabled the authorities to work with more reliable and competent data as 3D data (and maps) are proven to be more accurate and relatable in representing real world objects and scenarios. In South East Asia countries such as Malaysia and Singapore, urban development activities are dynamic causing city inventory information to change within short periods of time. The ability to detect and track the changes on building structures located in a city can avoid loss of revenue to the urban authorities as well as improving the efficiency of a city inventory in assisting the decision making process on management level. City inventory management system should be able to store and provide information especially related to the building structure that is needed by the local authority to manage a city efficiently. Current traditional methods utilized by most urban authorities in South East Asia region are inefficient and unable to cope with current complex scenarios in urban areas. This results in some valuable information for building management and information systems being discarded. This research presents a method to boost the efficiency of some aspect in city inventory for local authorities by introducing 3D change detection method on 3D city models. The proposed method is able to detect and analyse changes that occur on 3D buildings using the 3D segmentation approach. A hybrid semantic-geometric 3D segmentation method that deals with segmentation of a 3D building based on its semantic value and surface characteristics, fitted by one of the predefined primitives is presented. The technique is designed to retain existing semantic and geometric information and ultimately enrich it by adding more specialized information. For future work, the proposed method will be implemented as part of the auto-updating module that is able to detect any changes on 3D buildings such as storage and retrieval of semantic information of the changed structure, and automatically update the 3D city models for analysis.