This project, developing a versatile mobile range system based on the principle of Light Detection and Ranging (LIDAR), enables: - New data driven approaches in 3D reconstruction and modeling, pattern recognition, and data mining; - Quantitative studies in geology (e.g., analysis of hill slope movement and characterization of outcrops; - Archaeological recording of fragile sites with limited access (e.g., those in caves). The system consists of 3 LIDAR sensor heads, GPS and inertial measurement unit, high-resolution digital cameras, and processing software to generate high-resolution, accuracy colored point clouds that are geo-referenced. Unlike typical laser scanning system, it supports both stationary and high-speed mobile scanning. For example, when mounted on a vehicle, it should be able to generate over 50 points per meter given a speed of 100km/hr. In comparison, the typical density from an airborne LIDAR system is only 1-2 points per meter. This level of density at high scanning speeds allows rapid 3D scanning of large scenes not possible with either stationary systems (too slow) or airborne systems (too sparse). Through relatively simple hardware modification and more sophisticated processing algorithms, the state of the art in scanning range, density, and operational conditions is expected to advance as follows: - Increasing the scanning range over 6 times in the mobile mode, from the original 100 m (specified by the LIDAR sensor vendor) to greater than 600 m. - Improving the point density over 10 times by applying the principle of compressing sensing and information from the high-resolution color image, without changing the hardware configuration. - Developing the ability to scan beyond line of sight with inexpensive self-calibrating reflective mirrors, and develop real-time vision and inertial-based navigation to allow GPS-free operations for mobile scan (critical for scanning in underground areas, where many archaeological sites reside). Plans also include building a large scale 3D cityscape database with over 30 billions of raw data samples to cover the entire Lexington metro area. The database will be shared to enable research beyond graphics and vision, including, but not limited to, data compression, transmission, visualization, index and retrieval, and computational geometry.
Broader Impacts: The instrument is expected to be the first university-owned mobile laser range sensing system, which might inspire and facilitate many exciting new research venues. The scanning and processing tasks will be carried out and documented by undergraduate students. The geological and archaeological studies will be used to attract local students, in particular those underrepresented from Appalachia. The system will have commercial values to provide on-demand scanning for applications ranging from construction, city planning, law enforcement, survey and mapping, to 3D imaging. Commercial users will be charged a fee to maintain operation and sustain the enabling data beyond the life of the award.
This is an equipment development project. The goal is to develop a high-resolution 3D scanning system that can be used in a variety of operational conditions. The main intellectual merit is that a fully functional 3D scanning system has been developed as planned. As shown in the attached image, it contains two 3D Laser range sensors, a high-resolution panoramic camera, and a highly-accurate GPS/IMU (Inertial measurement unit) system. Software has been developed to (a) control all the system components from a laptop PC; (b) record data, and (c) synchronize and register data from different sources. While this is mainly a development project, new algorithms about sensor fusion and 3D modeling have been published in top conferences in both computer graphics and computer vision. Sample data captured by this system, which are over one gigabytes in size, has been shared online for research purpose. The broad impact from this project is that it provides a quick and accurate way to obtain 3D data at a large scale. The scanning system has been used by a number of cross-disciplinary projects across the UKy campus, enabling several new grants in Department of Transportation, Computer Science,and Geology. It has contributed in part to attract over $2 million of external funding from both private and public funding sources.
