CN108459319A - A kind of quick scanning system of vehicle running region Terrain Elevation - Google Patents

Abstract

The invention discloses a kind of quick scanning systems of vehicle running region Terrain Elevation, including four millimetre-wave radars, mobile lidar, five points cloud processing devices, display module and master controller；Four millimetre-wave radars are respectively placed in four corner locations of vehicle chassis, scan the roadblock and hill start of vehicle traveling direction corresponding region respectively；Mobile lidar is positioned over right over vehicle at 50 positions 80cm；Four millimetre-wave radars and mobile lidar carry out communication by Ethernet with respective points cloud processing device respectively and connect, and points cloud processing device carries out level-one point cloud bulk processing to millimetre-wave radar and mobile lidar initial data point cloud；Master controller is connected by Ethernet and the communication of each point cloud processor, and master controller obtains the point cloud data after points cloud processing device level-one bulk processing by Ethernet and carries out two level precision processing to laser radar corresponding region point cloud according to millimetre-wave radar information；Master controller is connected with display.The present invention have many advantages, such as stability it is high, can all weather operations.

Description

A fast scanning system for terrain height in vehicle driving area

technical field

The invention relates to the field of vehicle engineering detection, in particular to a fast scanning system for terrain heights in vehicle driving areas

Background technique

At present, vehicle active suspension technology solves the problem of vehicle ride comfort, but it has a certain lag in making adjustments and cannot make quick adjustments. In addition, although millimeter-wave radar alone has good detection capabilities in dust, smoke, rain and snow conditions in vehicle assisted driving, it has a great lack of acquisition accuracy. The acquisition accuracy of a single lidar is high, but its ability to adapt to severe weather is too poor. There are also some countries that combine millimeter wave and lidar, but the method of processing data is too slow, which greatly reduces the speed of data collection.

Contents of the invention

The purpose of the present invention is to provide a highly stable and all-weather working terrain height rapid scanning system in the vehicle driving area.

In order to achieve the above object, the following technical solutions are adopted: the system of the present invention includes four millimeter-wave radars, vehicle-mounted laser radars, five point cloud processors, a display module and a general controller; the four millimeter-wave radars are respectively placed on the vehicle The four corner positions of the chassis scan the roadblocks and slopes in the corresponding area of the vehicle's traveling direction; the vehicle-mounted laser radar is placed at a position 50-80cm directly above the vehicle and realizes 360°rotational scanning to collect terrain information around the vehicle; four The millimeter-wave radar and the vehicle-mounted lidar communicate with their respective point cloud processors through Ethernet, and the point cloud processor performs a first-level point cloud rough processing on the initial data point cloud of the millimeter-wave radar and vehicle-mounted lidar; The Ethernet communicates with each point cloud processor, and the general controller obtains the point cloud data after the first-level rough processing of the point cloud processor through the Ethernet, and performs two-level fine processing on the point cloud of the corresponding area of the lidar according to the millimeter-wave radar information; The master controller is connected with the monitor, and uploads the road elevation change status to the monitor.

Further, the first-level point cloud rough processing of the on-board lidar includes point cloud filtering, point cloud hole interpolation, point cloud compression, and converting the original data point cloud into a simplified data point cloud.

Further, the first-level point cloud rough processing of the millimeter-wave radar includes point cloud filtering and point cloud feature extraction, roughly simulates and calculates the terrain features of the vehicle travel area, and calculates the terrain change curve and prominent roadblocks in the vehicle travel area size, coordinates.

Further, the point cloud filtering adopts a combination of Gaussian filtering and mean curvature flow filtering, and the two kinds of filtering respectively filter out ordered point clutter and scattered point clutter.

Further, the method based on surface fitting is adopted to interpolate the point cloud data.

Further, the simplified point cloud data is compressed and exported with the region merging method.

Further, the second level of fine processing is to perform point cloud registration based on the point cloud data of the wheel's forward line and the area that the vehicle will pass through in the future, and change the point clouds in five different coordinate systems to the same coordinate system to improve measurement accuracy.

Compared with prior art, the present invention has following advantage:

1. The combination of millimeter-wave radar and laser radar can adapt to any bad weather and improve the reliability of the system.

2. There are two levels of point cloud data processing. The millimeter-wave radar data is used as the basis to eliminate the lidar data, and only the point cloud of the driving area is retained, which greatly reduces the number of redundant point clouds and improves the scanning efficiency of the system.

3. The control part of the present invention is composed of 5 point cloud processors and a general controller, which are connected together through Ethernet, and each point cloud controller is divided into tasks, and the general controller fuses the precise data, which once again improves the efficiency of data processing. speed.

Description of drawings

Fig. 1 is the overall layout structural diagram of the present invention.

Fig. 2 is a working flow diagram of the fast scanning method of the present invention.

Detailed ways

The present invention will be further described below in conjunction with accompanying drawing:

As shown in Figure 1, the system of the present invention includes four millimeter-wave radars, vehicle-mounted laser radars, five point cloud processors, a display module and a general controller; the four millimeter-wave radars are respectively placed on the four sides of the vehicle chassis Angle position, to scan the roadblocks and slopes in the corresponding area of the vehicle's traveling direction; the vehicle-mounted laser radar is placed at a position 50-80cm directly above the vehicle and realizes 360°rotational scanning to collect terrain information around the vehicle; four millimeter-wave radars and vehicle-mounted LiDAR communicates with its respective point cloud processor through Ethernet, and the point cloud processor performs a first-level point cloud rough processing on the initial data point cloud of millimeter-wave radar and vehicle-mounted LiDAR; the general controller communicates with each point cloud through Ethernet. Cloud processor communication connection, the general controller obtains the point cloud data after the first-level rough processing of the point cloud processor through Ethernet, and performs two-level fine processing on the point cloud of the corresponding area of the lidar according to the millimeter-wave radar information; the general controller and the display Connected to upload the road elevation change status to the display.

As shown in Figure 2, the specific workflow of the system is as follows: the data point cloud collected by the vehicle-mounted lidar and the millimeter-wave radar is transmitted to the corresponding point cloud processor through Ethernet, and the initial data point cloud of the radar is processed by the point cloud processor. Perform first-level point cloud rough processing. The first-level point cloud processing of lidar includes point cloud filtering, point cloud interpolation, point cloud compression, and converting the original data point cloud into a simplified data point cloud. The first-level point cloud processing of millimeter-wave radar includes point cloud filtering and point cloud feature extraction, roughly simulates and calculates the terrain features of the vehicle travel area, and calculates the terrain change curve of the vehicle travel area and the size and coordinates of prominent roadblocks. For point cloud filtering, the present invention adopts the combination of Gaussian filtering and average curvature flow filtering. The two kinds of filtering are to filter out ordered point clutter and scattered point clutter respectively. Useless data filtering.

Due to the limitations of the photoelectric sensor itself, the collected point cloud often includes various parts that cannot be measured. Therefore, the present invention aims at this problem and adopts a method based on surface fitting to interpolate the point cloud data to ensure the integrity of the data.

Finally, the simplified point cloud data is compressed and exported by the region merging method, which reduces the amount of data and improves the efficiency of the system while ensuring that the ground elevation data is not lost.

The general controller communicates with each point cloud processor through Ethernet to obtain the compressed and simplified data point cloud matrix. According to the change curve and roadblocks detected by the millimeter wave radar, the general controller finely processes the lidar and its corresponding area data , and delete the data in other areas at the same time, and register the point cloud data in the vehicle travel area through the roadblock signs and fluctuations, and the point cloud data in five different coordinate systems P and Q satisfy the same rigid body (R, T) Convert to the same horizontal coordinate system (R rotation matrix, T translation matrix), which can ensure the accuracy of the data, reduce the data point cloud, and greatly improve the scanning speed.

From the point cloud data under the same coordinate system, the elevation parameter value of the ground is extracted. On the one hand, a time series signal based on the terrain height in front of the vehicle is generated and uploaded to the display module for display. On the other hand, the vehicle suspension control amount is generated based on the terrain elevation change in front of the vehicle It adapts to the undulations of the terrain, ensuring smooth operation.

The above-mentioned embodiments are only descriptions of preferred implementations of the present invention, and are not intended to limit the scope of the present invention. All such modifications and improvements should fall within the scope of protection defined by the claims of the present invention.

Claims (7)

1. A fast scanning system for terrain height in vehicle driving area, characterized in that: the system includes four millimeter-wave radars, vehicle-mounted laser radar, five point cloud processors, display modules and general controllers; four millimeter-wave radars Place them on the four corners of the vehicle chassis to scan the roadblocks and slopes in the corresponding area in the direction of the vehicle; the vehicle-mounted laser radar is placed at a position 50-80cm directly above the vehicle and realizes 360°rotational scanning to collect the terrain around the vehicle Information; the four millimeter-wave radars and the vehicle-mounted lidar communicate with their respective point cloud processors through Ethernet, and the point-cloud processor performs a first-level point cloud rough processing on the initial data point cloud of the millimeter-wave radar and vehicle-mounted lidar; The general controller communicates with each point cloud processor through Ethernet. The general controller obtains the point cloud data after the first-level rough processing of the point cloud processor through Ethernet, and performs secondary processing on the point cloud of the corresponding area of the lidar according to the millimeter-wave radar information. Level fine processing; the main controller is connected with the monitor, and uploads the road elevation change status to the monitor. 2. A fast scanning system for terrain height in a vehicle driving area according to claim 1, characterized in that: the first-level point cloud rough processing of the vehicle-mounted laser radar includes point cloud filtering, point cloud hole interpolation, and point cloud compression. The raw data point cloud is transformed into a reduced data point cloud. 3. A fast scanning system for terrain height in a vehicle driving area according to claim 1, characterized in that: the first-level point cloud rough processing of the millimeter-wave radar includes point cloud filtering and point cloud feature extraction, and the vehicle driving area The terrain features are rough simulated and calculated, and the terrain change curve and the size and coordinates of prominent roadblocks are calculated in the area where the vehicle travels. 4. according to claim 2 or 3 described a kind of terrain height rapid scanning system of vehicle driving area, it is characterized in that: described point cloud filtering, adopts the mode that Gaussian filtering and average curvature flow filtering combine, two kinds of filtering respectively to Ordered point clutter and scattered point clutter are filtered out. 5. A fast scanning system for terrain height of vehicle driving area according to claim 1, characterized in that: the point cloud data is interpolated by adopting a method based on surface fitting. 6. A fast scanning system for the terrain height of the vehicle driving area according to claim 1, characterized in that: the simplified point cloud data is compressed and exported by the area merging method. 7. A fast scanning system for terrain height in a vehicle driving area according to claim 1, characterized in that: the secondary finishing process is to carry out point cloud registration according to the point cloud data of the straight line of the wheel and the area that the vehicle will pass through in the future, and The point clouds in five different coordinate systems are switched to the same coordinate system to improve the measurement accuracy.