Integrating lidar into the Community Sensor Model construct

There is an ever increasing demand for geospatial data and in recent years this demand has expanded from the 2D to the 3D realm. Certain applications using this data require the ability to rigorously propagate and predict the errors associated with these geospatial datasets. This includes errors associated with initial data collection, but also includes refinements made in post processing. To meet the community´s needs, the National Geospatial-Intelligence Agency (NGA) and the Air Force Aeronautical Systems Center initiated a program called the Community Sensor Model Working Group (CSMWG). The CSMWG maintains an application program interface (API) that standardizes the means for sensor exploitation tools to access physical sensor model methods without insight into the detailed and often proprietary mathematics describing the sensing system. Key methods relate a sensed object to the sensor parameters, assign a measure of precision to the transformation, and provide a means to adjust the parameters for data registration. To date Light Detection and Ranging (lidar) has not been fully integrated into CSM, even though lidar has become a primary collector for 3D data. As the quantity of lidar datasets has increased, the need to merge multiple lidar datasets and fuse them with other data has become prevalent. Maximizing the utility of lidar and fused products requires adjusting the lidar and/or the other data to improve the relative registration and possibly the absolute accuracy. Performing these adjustments in a rigorous manner requires precise physical sensor models. A Universal Lidar Error Model (ULEM) has been proposed to meet this need. This paper provides a high level overview of CSM and associated CSM functionality. ULEM is then presented as a method to meet CSM compliancy for lidar models; providing benefits during rigorous error propagation, block adjustment of lidar data and the fusion of lidar with other data types. Two potential ULEM implementation methods a- e presented, along with discussion on the potential necessity for multiple implementations. Sample results are presented from recent implementation of ULEM on existing systems. The paper concludes with a discussion of the current status and future direction of ULEM.