阅读说明：本技术 一种城市数据采集系统 (City data acquisition system ) 是由 何伟 张文强 于 2021-07-30 设计创作，主要内容包括：本发明提供了一种城市数据采集系统,包括：移动载体；多平台移动测量模块,设置在移动载体中,包括用于获取激光点云数据的三维激光扫描仪、用于获取影像数据的CCD摄像机、用于配合获取组合定位定姿数据的定位模块以及微电脑控制器；标定板,设置在移动载体中,用于对三维激光扫描仪的坐标系和CCD摄像机的坐标系的空间一致性进行约束,并通过标定算法计算得到三维激光扫描仪和CCD摄像机的标定参数,其中,定位模块包括用于获取定位数据的北斗导航单元和用于获取姿态数据的惯性测量单元,微电脑控制器中还具有立体显示模块进行立体显示,微电脑控制器中还具有车载移动测量模块优化来得到最优标定参数。(The invention provides a city data acquisition system, comprising: moving the carrier; the multi-platform mobile measurement module is arranged in the mobile carrier and comprises a three-dimensional laser scanner for acquiring laser point cloud data, a CCD camera for acquiring image data, a positioning module for acquiring combined positioning and attitude determination data in a matching manner and a microcomputer controller; the calibration board is arranged in the mobile carrier and used for constraining the space consistency of a coordinate system of the three-dimensional laser scanner and a coordinate system of the CCD camera and calculating calibration parameters of the three-dimensional laser scanner and the CCD camera through a calibration algorithm, wherein the positioning module comprises a Beidou navigation unit used for acquiring positioning data and an inertia measurement unit used for acquiring attitude data, the microcomputer controller is also provided with a stereo display module for stereo display, and the microcomputer controller is also provided with a vehicle-mounted mobile measurement module for optimizing to obtain optimal calibration parameters.)

1. A city data collection system, comprising:

moving the carrier;

the multi-platform mobile measurement module is arranged in the mobile carrier and comprises a three-dimensional laser scanner for acquiring laser point cloud data, a CCD camera for acquiring image data, a positioning module for acquiring combined positioning and attitude determination data in a matching manner and a microcomputer controller;

a calibration plate arranged in the mobile carrier and used for restricting the space consistency of the coordinate system of the three-dimensional laser scanner and the coordinate system of the CCD camera and calculating calibration parameters of the three-dimensional laser scanner and the CCD camera by a calibration algorithm,

wherein the positioning module comprises a Beidou navigation unit for acquiring positioning data and an inertial measurement unit for acquiring attitude data,

the microcomputer controller is also provided with a stereo display module for carrying out high-precision three-dimensional digital modeling to carry out stereo display according to the laser point cloud data and the image data,

the microcomputer controller is also provided with a vehicle-mounted mobile measurement module which is used for enabling the three-dimensional laser scanner to conduct multiple scanning on the same ground object to be overlapped to serve as a constraint condition, optimizing the calibration parameters by using an LM optimization algorithm, optimizing the calibration parameters by enabling errors among multiple scanning identical points to be minimum, and acquiring the laser point cloud data and the image data after the three-dimensional laser scanner and the CCD camera are calibrated according to the optimal calibration parameters.

2. The city data collection system of claim 1, wherein:

wherein, the multi-platform mobile measurement module also comprises a receiver, a sensor main control circuit board, a video distribution and gating circuit board and an image acquisition card,

the sensor main control circuit board and the video distribution and gating circuit board are positioned in the microcomputer controller, the receiver is used for receiving signal data of the Beidou navigation unit, the image acquisition card is used for receiving image data of the CCD camera,

the sensor main control circuit board respectively performs data interaction with the receiver, the three-dimensional laser scanner and the video distribution and gating circuit board,

the video distribution and gating circuit board is connected with the image acquisition card.

3. The city data collection system of claim 1, wherein:

and the three-dimensional laser scanner fuses original laser scanning data with external calibration parameters of the three-dimensional laser scanner and the combined positioning and attitude determination data and then reconstructs the data to obtain three-dimensional laser point cloud data.

4. The city data collection system of claim 1, wherein:

the microcomputer controller is also provided with an image segmentation module, the image segmentation module is used for carrying out pixel-level classification on different types of targets, carrying out 3D modeling on complex conditions of a plurality of overlapped objects and different backgrounds through the image data acquired by the CCD camera, and determining the boundary, the type and the relationship among the objects.

5. The city data collection system of claim 1, wherein:

the microcomputer controller is also provided with a data fusion module, and the data fusion module is used for generating a high-resolution characteristic diagram according to the laser point cloud data and the image data.

Technical Field

The invention belongs to the field of data acquisition, and particularly relates to a city data acquisition system.

Background

At present, the three-dimensional panoramic technology captures image information of a whole scene through a professional camera, uses software to combine pictures, and uses a special player to play, namely, a plane photograph and a computer are changed into a 360-degree full-view scene for virtual reality browsing. The two-dimensional plane graph is simulated into a real three-dimensional space and presented to an observer, various functions of manipulating images are provided for the observer, the images can be amplified and reduced, and the scene can be watched by moving in various directions, so that the effect of simulating and reproducing the real environment of the scene is achieved. The advantages are that: the sense of reality is strong; the generation is convenient; the manufacturing period is short; the file is small, the transmission is convenient, the network use is suitable, the release format is various, and the like. However, the existing three-dimensional panoramic data acquisition method is long in time consumption and high in cost, and is more complex particularly under the condition of panoramic data acquisition in urban complex environment space.

Disclosure of Invention

The present invention is made to solve the above problems, and an object of the present invention is to provide an urban data collection system.

The invention provides a city data acquisition system, which is characterized by comprising the following components:

moving the carrier; the multi-platform mobile measurement module is arranged in the mobile carrier and comprises a three-dimensional laser scanner for acquiring laser point cloud data, a CCD camera for acquiring image data, a positioning module for acquiring combined positioning and attitude determination data in a matching manner and a microcomputer controller; the calibration board is arranged in the mobile carrier and used for constraining the space consistency of a coordinate system of the three-dimensional laser scanner and a coordinate system of the CCD camera and obtaining calibration parameters of the three-dimensional laser scanner and the CCD camera through calibration algorithm calculation, wherein the positioning module comprises a Beidou navigation unit used for obtaining positioning data and an inertial measurement unit used for obtaining attitude data, the microcomputer controller is also provided with a stereo display module used for carrying out high-precision three-dimensional digital modeling according to laser point cloud data and image data to carry out stereo display, the microcomputer controller is also provided with a vehicle-mounted mobile measurement module used for superposing the laser point clouds of multiple scans of the same ground object by the three-dimensional laser scanner as constraint conditions, optimizing the calibration parameters by using LM optimization algorithm, and optimizing the calibration parameters by minimizing the error between multiple times of scanning the same-name point to obtain optimal calibration parameters, and calibrating the three-dimensional laser scanner and the CCD camera according to the optimal calibration parameters and then acquiring urban data.

The city data acquisition system provided by the invention can also have the following characteristics: the multi-platform mobile measurement module further comprises a receiver, a sensor main control circuit board, a video distribution and gating circuit board and an image acquisition card, wherein the sensor main control circuit board and the video distribution and gating circuit board are located in the microcomputer controller, the receiver is used for receiving signal data of the Beidou navigation unit, the image acquisition card is used for receiving image data of the CCD camera, the sensor main control circuit board is respectively in data interaction with the receiver, the three-dimensional laser scanner and the video distribution and gating circuit board, and the video distribution and gating circuit board is connected with the image acquisition card.

The city data acquisition system provided by the invention can also have the following characteristics: the three-dimensional laser scanner fuses original laser scanning data with external calibration parameters of the three-dimensional laser scanner and combines, positions and positions the attitude data, and then the three-dimensional laser scanning data is reconstructed to obtain three-dimensional laser point cloud data.

The city data acquisition system provided by the invention can also have the following characteristics: the microcomputer controller is also provided with an image segmentation module, the image segmentation module is used for carrying out pixel-level classification on different types of targets, carrying out 3D modeling on complex conditions of a plurality of overlapped objects and different backgrounds through image data acquired by the CCD camera, and determining the boundary, the type and the relation among the objects.

The city data acquisition system provided by the invention can also have the following characteristics: the microcomputer controller is also provided with a data fusion module, and the data fusion module is used for generating a high-resolution characteristic diagram according to the laser point cloud data and the image data.

Action and Effect of the invention

According to the urban data acquisition system, four sets of core sensors, namely a Beidou navigation unit, a three-dimensional laser scanner, a CCD camera and an inertia measurement unit, are integrated in the multi-platform mobile measurement module, so that the high-efficiency fusion of multi-source data can be realized, and the three-dimensional panoramic data of the urban complex environment space can be efficiently acquired; because the mobile carrier is arranged to carry the multi-platform mobile measurement module, the mobile device can conveniently move in a city to acquire data, and the upgrading of data acquisition equipment and the innovation of technical means are realized; the calibration board is arranged to calculate to obtain the calibration parameters, and the calibration parameters are optimized to obtain the optimal calibration parameters, so that the calibration accuracy can be further ensured, and the data acquisition accuracy is effectively improved. Moreover, the invention realizes the localization of the space panoramic data acquisition equipment, fills up domestic shortcuts, and promotes the application of the space panoramic data based on the Beidou in the construction of novel smart cities and digital twin cities.

Drawings

FIG. 1 is a block diagram of a city data collection system in an embodiment of the invention;

FIG. 2 is a schematic view of a scanning pattern of a three-dimensional laser scanner in an embodiment of the invention;

FIG. 3 is a block diagram of a multi-platform mobility measurement module according to an embodiment of the present invention;

fig. 4 is a design flowchart of a stereoscopic display module performing stereoscopic display through three-dimensional numerical modeling in an embodiment of the present invention.

Detailed Description

In order to make the technical means and functions of the present invention easy to understand, the present invention is specifically described below with reference to the embodiments and the accompanying drawings.

< example >

Fig. 1 is a block diagram of a city data collection system in an embodiment of the present invention.

As shown in fig. 1, a city data collection system 100 of the present embodiment includes a mobile carrier 10, a multi-platform mobile measurement module 20, and a calibration board 30.

The movable carrier 10 includes a human machine, an unmanned aerial vehicle, a manned vehicle, an unmanned vehicle, a manned ship, an unmanned ship, a backpack and other carriers, and is used for moving.

The multi-platform mobile measurement module 20 is disposed in the mobile carrier 10, and includes a three-dimensional laser scanner for acquiring laser point cloud data, a CCD camera for acquiring image data, a positioning module for acquiring combined positioning and pose-fixing data in a coordinated manner, and a microcomputer controller.

The positioning module comprises a Beidou navigation unit for acquiring positioning data and an inertial measurement unit for acquiring attitude data.

In this embodiment, the beidou navigation unit is a high-precision measurement type, and the technical index design includes signal tracking ability, positioning accuracy, time service accuracy, speed measurement accuracy, data update rate, first positioning time, and signal recapture time.

The three-dimensional laser scanner is a small and exquisite light model with high emission frequency, and the technical index design comprises the whole emission frequency, the single machine weight, the laser grade, the maximum scanning measuring range, the scanning range, the distance measuring precision and the scanning frequency.

Fig. 2 is a schematic view of a scanning pattern of the three-dimensional laser scanner in the embodiment of the present invention.

As shown in fig. 2, the three-dimensional laser scanner is in a line scanning mode, and section scanning is performed during the traveling process of the moving carrier, for example, a thick solid line L in fig. 2 represents an effective scanning boundary of the three-dimensional laser scanner, an area between two thick solid lines represents an effective scanning area, a thin solid line H represents a laser beam, a broken line P represents a surface of a ground object including a road surface and an outer contour of a building, a point M represents a laser scanning point, and as the moving carrier advances, a continuous section constitutes a three-dimensional sampling expression of the ground object. In fig. 2, a black solid line indicates an outline of a feature, a straight line with an arrow indicates a track and a direction, a dot indicates a laser spot, the density of the laser spot is higher in an area close to the laser scanner than in an area far from the laser scanner, and a solid line indicates one scanning line.

In this embodiment, the measurement of each sensor in the multi-platform mobile measurement module 20 is completed under the coordinate system of the sensor itself, the original scanning data loaded to the mobile carrier by the three-dimensional laser scanner is based on the scanner coordinate system defined by the three-dimensional laser scanner, and the scanning is performed in one plane in the scanning process relative to the three-dimensional laser scanner itself. Therefore, three-dimensional reconstruction of three-dimensional laser requires scanning laser data with original data and converting the laser data into laser data under a world coordinate system, and the three-dimensional laser scanner reconstructs original laser scanning data into three-dimensional laser point cloud data after fusing external calibration parameters of the three-dimensional laser scanner and combining, positioning and attitude determination data according to a laser scanner positioning equation.

Fig. 3 is a schematic diagram of an architecture of a multi-platform mobile measurement module according to an embodiment of the present invention.

As shown in fig. 3, the multi-platform mobile measurement module 20 further includes a receiver, a sensor main control circuit board, a video distribution and gating circuit board and an image acquisition card,

the sensor main control circuit board and the video distribution and gating circuit board are positioned in the microcomputer controller,

the receiver is used for receiving signal data of the Beidou navigation unit, the image acquisition card is used for receiving image data of the CCD camera,

the sensor main control circuit board is respectively in data interaction with the receiver, the three-dimensional laser scanner and the video distribution and gating circuit board, and the video distribution and gating circuit board is connected with the image acquisition card.

The multi-platform mobile measurement module 20 can realize real-time, active and complete three-dimensional space panoramic data acquisition by effectively integrating various sensors such as a Beidou navigation unit, a three-dimensional laser scanner, an inertia measurement unit, a CCD camera and a microcomputer controller. And obtaining high-precision ground object surface point cloud and high-definition images through Beidou/IMU integrated navigation resolving, time synchronization processing, fusion of laser data and other sensor data and three-dimensional space calculation. The multi-platform mobile measurement module 20 can be multipurpose, is loaded on mobile carriers such as a man-machine, an unmanned machine, a man-vehicle, an unmanned vehicle, a man-ship, an unmanned ship and a backpack, realizes various operation modes such as an airborne mode, a vehicle-mounted mode, a ship-mounted mode and a backpack mode, and can collect mass point cloud data and high-definition image data with high precision and high efficiency.

The microcomputer controller is also provided with a three-dimensional display module for performing high-precision three-dimensional digital modeling according to the laser point cloud data and the image data to perform three-dimensional display.

In this embodiment, the stereo display module uses the FPGA as an image processing core, and performs matching and stereo display in real time according to laser point cloud data and image data acquired by collecting a three-dimensional scene in a city by using abundant resources and high-speed computing power inside the FPGA, and performs ranging on a significant target in the scene.

Fig. 4 is a design flowchart of a stereoscopic display module performing stereoscopic display through three-dimensional numerical modeling in an embodiment of the present invention.

As shown in fig. 4, the stereoscopic display module mainly includes PAL video image acquisition and display, implementation of Census transform of a local stereoscopic matching algorithm, VGA display, salient object extraction and rangefinder, and stereoscopic display.

The microcomputer controller is also provided with a vehicle-mounted mobile measurement module which is used for enabling the three-dimensional laser scanner to conduct repeated scanning on the same ground object to be overlapped to serve as a constraint condition, optimizing calibration parameters by using an LM (Linear motion estimation) optimization algorithm, optimizing the calibration parameters by enabling errors among the repeated scanning same-name points to be minimum to obtain optimal calibration parameters, and acquiring urban data after the three-dimensional laser scanner and the CCD camera are calibrated according to the optimal calibration parameters.

The microcomputer controller is also provided with an image segmentation module which is used for carrying out pixel-level classification on different types of targets, carrying out 3D modeling on complex conditions of a plurality of overlapped objects and different backgrounds through image data acquired by the CCD camera, and determining the boundary, the type and the relationship among the objects.

In this embodiment, for the acquisition of spatial panoramic data of a complex environment in a city, it is not sufficient to perform 3D modeling on the surrounding environment only by using sensors, and it is necessary to understand the surrounding environment intelligently and further process and analyze the detected objects, for example, to obtain the category, number, and volume (boundary) of the detected objects. The image segmentation module of the invention classifies different types of targets at pixel level by an example segmentation technology, performs 3D modeling on the complex situations of a plurality of overlapped objects and different backgrounds by using a visual sensor, and determines the boundary, the type and the relationship among the objects, which are specifically as follows:

the instance segmentation has the characteristics of semantic segmentation, and needs to be classified on a pixel level, and also has a part of characteristics of target detection, namely different instances need to be positioned even if the instances are of the same type. The invention uses the thought of graph display technology Render for reference, and utilizes a non-uniform sampling mode to determine the sawtooth phenomenon when the scale mode changes under the condition of improving the resolution. By determining points on the boundary and judging the attribution of the points, the image segmentation of the object edge is optimized, so that the object edge has better performance in the edge part of the object which is difficult to segment; in addition, when performing 3D modeling on the complex situation of a plurality of overlapped objects and different backgrounds and determining the boundary, the type and the relationship among the objects, target detection is generally performed first, then binary image segmentation is performed on each target, and because the target detection index and the classification score are not well correlated, the image segmentation module calibrates the deviation between the mask quality and the target frame index by predicting the overall mask segmentation score, and improves the example segmentation performance. In addition, precision of a target detection frame is improved by introducing precision RoI Pooling.

The microcomputer controller is also provided with a data fusion module which is used for generating a high-resolution characteristic diagram according to the laser point cloud data and the image data.

In this embodiment, the data fusion module is used for generating a high-resolution feature map from the radar point cloud and the RGB image, and has a high detection performance for the object.

The method comprises the following specific steps: although radar and camera are the mainstream target detection sensors at present, with the gradual development of spatial data acquisition to the direction of multisource (multi-platform, multi-sensor, multi-angle) and high resolution (high spatial resolution, high temporal resolution, high spectral resolution, high radiation resolution), the advantage of radar point cloud + camera image shows gradually, and the integration of these two can greatly promote the precision: the radar can directly acquire the three-dimensional coordinates of a target, has unique advantages in the aspect of acquiring vertical structure information of buildings and vegetation, but lacks texture and spectrum information which are significant for target identification, and the resolution of radar point cloud is reduced along with the increase of detection distance; camera images can acquire rich texture information and spectral information, but are sensitive to light and shadows, and cannot provide accurate and sufficient position information when positioning an object in 3D space, which generally results in poor real-time performance and robustness. The combination of radar and camera can realize complementary advantages well.

The data fusion module of the invention fully utilizes the point cloud depth information and the image semantic information, simultaneously feeds in the point cloud and the image, and generates reliable 3D object and distance suggestion for the radar point cloud and the RGB image based on the high-resolution characteristic diagram. Typically, a 3D mesh map is built on the radar point cloud before using the 3D information, and then the point cloud is processed on the mesh map. Although the 3D mesh representation retains most of the original information of the point cloud, it typically requires more complex calculations for subsequent processing. The data fusion module of the invention projects the radar point cloud into the 2D space, reduces the calculated amount, and fuses the data of a plurality of sensors by utilizing a probability data association algorithm. By integrating multiple sensors to replace the recognition results of the representative features, an optimal global decision can be made due to the independent detection and classification based on each sensor.

In addition, the data fusion module of the invention provides a novel feature extractor structure, and the structural design can generate a high-resolution feature map from the radar point cloud and the RGB image, and has higher detection performance for the prediction of an object; secondly, fusing the high-resolution feature maps to generate reliable 3D target suggestions; furthermore, ROI (region of interest) merging of each feature map is used to obtain equal length feature vectors, and the ROI merged feature maps are fused using element-wise mean operations. Key points in the fusion process are mainly temporal synchronization (time stamp resolution) and spatial synchronization (calibration resolution), which involve multiple processes including target association + target filtering + target tracking, etc.

The calibration plate 30 is disposed in the mobile carrier 10, and is used for constraining spatial consistency of a coordinate system of the three-dimensional laser scanner and a coordinate system of the CCD camera, and obtaining calibration parameters of the three-dimensional laser scanner and the CCD camera through a calibration algorithm.

In this embodiment, when the calibration board 30 is used for a low-cost novel solid-state laser radar, a laser radar point cloud registration algorithm based on different scanning modes is provided, and a motion model construction distortion correction algorithm is provided for the point cloud distortion problem caused by motion in the registration process.

Effects and effects of the embodiments

According to the urban data acquisition system related by the embodiment, as four sets of core sensors, namely a Beidou navigation unit, a three-dimensional laser scanner, a CCD camera and an inertia measurement unit, are integrated in the multi-platform mobile measurement module, the high-efficiency fusion of multi-source data can be realized, and the three-dimensional panoramic data of the urban complex environment space can be efficiently acquired; because the mobile carrier is arranged to carry the multi-platform mobile measurement module, the mobile measurement module can be conveniently moved in a city to acquire data, and the upgrading of data acquisition equipment and the innovation of technical means are realized; because the calibration board is arranged to calculate to obtain the calibration parameters and optimize the calibration parameters to obtain the optimal calibration parameters, the calibration accuracy can be further ensured and the data acquisition accuracy is effectively improved. Moreover, the embodiment realizes localization of space panoramic data acquisition equipment, fills up domestic short boards, and promotes application of the space panoramic data based on the Beidou in the construction of novel smart cities and digital twin cities.

The above embodiments are preferred examples of the present invention, and are not intended to limit the scope of the present invention.