期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2010
卷号：XXXVIII - 4/W13
出版社：Copernicus Publications
摘要：Road accidents have a definite impact in terms of lives, welfare and traffic congestion, especially with increase in mobility and number of overall vehicles. Prevention policies can greatly benefit from data analysis of positions and accident characteristics. This work presents a practical integration of open source software for the implementation of an instrument to insert, store, automatically analyze accident data, and consult results. As different public bodies survey accident data, a collaborative environment is necessary, and a web-based solution is ideal for permitting multi-user access and data insertion. PostGIS and UMN Mapserver with PHP server- side scripting are a classic solution for this type of instruments; this note will discuss how to organize these elements in order to build a portal for optimal data insertion, automatic statistical analysis and visualization of final result. Road center axes are processed to extract homogeneous segments which are then connected with a one-to-many spatial relationship of "closest-point-to- segment" road accident georeferenced positions. Accident points are also connected with a series of tables reporting accident- specific information. From this data an automatic procedure can extract specific indices to represent risk and danger, such as severity index, mortality index, vulnerability and other. Not only can this information be viewed as a thematic layer for fast interpretation, but change over time can be also analyzed when datasets are present over multiple years. With this information public decision makers can assess efficiency of past polices and decide future intervention
