1 Overview of 3D laser mapping lidar technology
1.1 The working principle of 3D lidar
Three-dimensional Lidar (LIDAR) is the abbreviation for laser detection and ranging system. The laser is used as the emitting light source, and the active alignment equipment adopts photoelectric detection technology. Lidar is an advanced detection method that combines laser technology and modern photoelectric detection technology. The transmitting system is a variety of lasers; the receiving system adopts a combination of telescopes and various forms of photodetectors, such as photomultiplier tubes, semiconductor photodiodes, avalanche photodiodes, infrared and visible light multi-detection devices.
When the vehicle-mounted three-dimensional lidar system is working, each sensor works under a unified reservation aimed at the terminal. GPS provides the coordinate position of the vehicle, and the inertial navigation system measures the attitude change of the vehicle in real time; the laser scanner emits laser pulses at frequencies above 10,000 Hz. The distance of the target is measured by measuring the flight time of the laser between the scanner and the target, the detection area is irradiated point by point through the scanning of the scanning mechanism, and the emission direction of each laser is measured by the angle measuring sensor, forming a line at the origin of the laser scanning coordinate Three-dimensional target information (distance, azimuth, pitch angle); CCD camera obtains the attribute information of the target in the detection area, and after each sensor is uniformly calibrated in position and time, through calculation, the spatial position of the target in the detection area (X, Y, Z) and attribute (texture) information. Through the movement of the carrying vehicle, the system achieves continuous information acquisition of targets on both sides of the road.
1.2 Application status of 3D lidar in the field of surveying and mapping
The 3D laser mapping lidar is a system that integrates three technologies: laser, global positioning system (GPS) and inertial navigation system (INS) to obtain data and generate accurate DEM. The combination of these three technologies can be divided into the currently maturing terrain LIDAR system for obtaining the ground digital elevation model (DEM) and the hydrological LIDAR system for obtaining the underwater DEM which has been matured. Locate the spot where the laser beam hits the object. The common feature of these two systems is the use of lasers for detection and measurement. The laser itself has a very precise ranging ability, and its ranging accuracy can reach a few centimeters. The accuracy of the LIDAR system is only a factor of the laser itself, and it also depends on the development of commercial GPS and IMU. Through LIDAR from mobile platforms (such as aircraft Above) Obtaining high-precision data has become possible and widely used.
At present, the use of 3D laser surveying and mapping lidar to form an airborne 3D laser lidar system has developed to the third generation, and the application is relatively mature. However, the development of vehicle-mounted 3D laser surveying and mapping lidar is relatively slow. Therefore, this article mainly focuses on the development of vehicle-mounted 3D laser surveying and mapping lidar. The key application technology is analyzed and discussed, in order to provide a series of technical foundation and guarantee for the development of a specific vehicle-mounted 3D laser mapping lidar system.
2 Analysis of key application technologies of vehicle-mounted 3D laser mapping lidar
2.1 Analysis of 3D data collection method
① Collection of terrain three-dimensional data. For the acquisition of terrain three-dimensional data, digital photography is usually used for measurement. However, because the ground surface is not smooth and continuous and uniform, the data segment will be broken during the collection process, and for densely vegetation and forest-covered areas , It is impossible to collect data in real time. In order to overcome this problem, in the actual three-dimensional digital measurement, it is decentralized measurement. Dispersed and batched measurement according to the layout points can realize real-time measurement of the ground surface, and the measurement accuracy can also meet the requirements of related applications.
② Collection of three-dimensional data of ground features. The collection of 3D data of ground objects is mainly realized through digital photography, and usually also realized by means of GIS satellite communication. It can obtain the three-dimensional spatial data of the features of the features in a comprehensive manner, and then transmit the three-dimensional spatial data of the features to the topographic mapping system through a dedicated transmitter to realize the spatial representation of the features. However, after the topographic map surveying and mapping system receives the three-dimensional data, it is also very important to recognize and interpret the three-dimensional data. Because there are inevitably some hidden points in the three-dimensional spatial data of many ground objects, the extraction and interpretation of the hidden points must be Preprocessing is required, and the true shape or position of the features can be calculated according to the front or rear intersection and related distance parameters, which provides a basic data basis for obtaining correct three-dimensional spatial data of the features.
2.2 Lidar point cloud and image registration fusion technology
Three-dimensional lidar data mainly includes laser point cloud, CCD image and position and attitude data of the vehicle. The three-dimensional laser cloud image contains the relative spatial position information of the target surface, but lacks the target's color, texture, and other attribute information. A single optical image is just the opposite. Its resolution is high and can truly record the texture and other details of the target surface. Cannot reflect the spatial position of the target surface. Therefore, the measurement data of different resolutions must be registered and fused according to the characteristics of the respective data of each sensor.
In order to quickly realize the registration and fusion of laser point cloud and image data in real time, the position registration method can be adopted. The specific steps are as follows:
First, preprocess the laser point cloud data. Complete the elimination of laser scanning point cloud electrical noise and background noise, and the homogenization of the scanning point cloud grid.
Secondly, perform distortion processing of image data to correct image distortion caused by optical lens. Finally, perform comprehensive calibration of 3D laser, GPS and IMU. Simultaneously establish solution methods and engineering practical algorithms to realize the spatial integration of multivariate data in a unified coordinate system.
2.3 High-speed storage technology for massive data
Since the biggest feature of vehicle-mounted 3D laser surveying and mapping lidar is its mobility, for vehicle-mounted laser systems, while realizing 3D mapping, a large amount of 3D cloud data needs to be processed quickly and stored in real time. However, as we all know, the data flow of 3D data is very large. If it cannot be processed and stored in real time, it may cause overflow or loss of 3D laser mapping data, which will affect the accuracy of 3D mapping. Therefore, it is necessary to solve the problem of fast storage of massive data in the vehicle-mounted 3D laser mapping lidar system.
In order to make the call of massive data efficient and convenient, the three-dimensional spatial data surveyed and mapped can be organized and managed in the following data organization methods:
The first is vector data, which is stored separately according to the map frame and scale; the second is raster data, which is regular grid data with a certain grid spacing, and its data organization is different from vector data.www.isurestar.net