| Over the past decade, with the development of society, economy and the progress of technology of surveying and mapping, great changes have taken place in the industry of surveying, mapping and geo-information. A rapid, economical and efficient way of geospatial data acquisition completely changed the traditional mode of operation and workflow in the industry, such as the application of navigation and positioning services, satellite remote sensing, unmanned aerial vehicle survey, and mobile mapping for land vehicle.With the advancement of the technologies, and more and more application that are widely used in digital cities, smart cities, geographical condition monitoring and other fields, as well as the continuous efforts with scholars in the field of surveying and mapping, mobile mapping system for land vehicle (VMMS), as a high-tech mapping instrument and equipment, has become one of the most dominant tools for data collection and is more and more important in the production. VMMS is a three-dimensional coordinate measurement and positioning system. And position, distance and angle can be measured by VMMS. Also, it has automation and digitization in the photography functions. In addition, the system integrates many technologies, such as GNSS unit, an inertial measurement unit (IMU), laser scanners, digital cameras, and automatic control and so on. Its aim is to collect multiple sets of information quickly, including spatial data, attribute data and real images in target areas and so on.Comprehensive accuracy of VMMS is always the most important factor in the actual application. In order to ensure its comprehensive accuracy, rigorous calibration methodologies are always desired and calibration must be carried out before data processing. Thus, calibration is one of the difficult points as well as a key point. Without calibration, accuracy is impossible. Under this context, a new system positioning model is firstly developed and system error sources are summarized in this paper. Furthermore, a detailed analysis of the influence of the error on system accuracy is given. Moreover, rigorous calibration approaches using both single calibration and combined calibration of VMMS are proposed, and the assessment of indexes and methods is also put forward on comprehensive accuracy. The main contents and innovations of this research are as follows:1. The components of VMMS were introduced in detail, as well as the working principle and theoretical basis of each part of VMMS. Then, the functions and roles of each part in the system were expounded. A detailed derivation of positioning model in the system was also given. And the key technologies of hardware integration were analyzed.2. The sources of errors and their processing methods were analyzed, as well as error types in VMMS related with system hardware integration and sensors. Each error model was deduced, and its influence on synthetic accuracy was quantitatively analyzed. System accuracy was theoretically estimated with positioning model, which provides references for selecting sensors while system integration.3. The single calibration of VMMS was proposed.1PPS timing accuracy was evaluated for GNSS receiver by rubidium clock with 1*10-11s and time counter with 5*10-11s. And precision evaluation for GNSS PPK was evaluated by short baseline and zero-baseline through GAMIT dynamic programming. IMU attitude angle precision was evaluated with GT580 and POSLV610. The angle precision of IMU can be evaluated more accurately with the two methods. Muti-tooth indexing table with 0.3" was used to evaluate angle precision for laser scanner. This method has high accuracy, good stability and quantity traceability. Digital camera was calibrated by angle measurement. The method is made by measuring incident angle of parallel rays and point coordinates on their corresponding pictures. And then inner orientation elements and distortion parameters were calculated based on principle of least square sum of the lens distortion. Results show that the corrected residual distortion is 0.2 pixels. This method is done indoor, besides, the calibration conditions and process is controllable and standard.4. The combined calibration of VMMS was proposed. Three methods that are manual method, method based on the control points and method based on feature plane were also proposed for combined calibration of laser scanner and IMU. The advantages and disadvantages of each approach were analyzed. The method based on the control points has highest accuracy with 0.001°, while the manual method has the lowest efficiency and the lowest accuracy. The combined calibration of digital camera and IMU was carried out by the space forward intersection and known feature points. The image coordinates were calculated with back calculation by point clouds coordinates. And RGB information of image points was assigned to point clouds for completing combination calibration of laser scanner and digital camera. Overlap accuracy of point clouds and images have reached 1-2 pixels.5. Permanent calibration field was established with homemade special symbols and reflections as feature points, which are convenient for point cloud feature extraction. The evaluation indicators for comprehensive accuracy of VMMS were put forward. The high precision three-dimensional coordinates and the length of characteristic lines of feature point clouds were obtained using the traditional method as a benchmark, with which the feature points and lines calculated by the system could be compared. Finally the comprehensive accuracy evaluating was completed. And the system accuracy in the actual engineering application was also assessed. The study showed the plane accuracy reached 0.028m, and elevation accuracy reached 0.019m for a calibrated VMMS, which tests the validity and efficiency of single calibration and combined calibration proposed in the paper. |
PDF Full Text Request