长测程地基激光雷达几何校正研究

发布时间：2018-10-08 19:34

【摘要】：激光雷达因为其能够快速获取目标物的三维空间信息以及具有全天候、精度高、费用低等优点,因此能迅速的在城市、交通、水利等领域中得到了广泛的应用,因此对于其精度的检校方面也吸引着许多学者的研究。现有的测距校正是基于对波形参数的校正,但由于波形参数校正存在着原始波形数据的封闭,而导致波峰探测和波形校正的方法的不明确。现有的波形参数提取是基于高斯混合模型对回波波形的参数提取,但这种模型假设存在着与波形数据的非线性、非正态分布形态和波形尾部抬升不一致性,导致波形参数提取的精度受限,而相对辐射校正方法无法直接的得到地物的反射特性。基于波形参数对测距校正的方法也没有考虑到仪器设备的衰减因素。此外,二维激光雷达是通过电流变化来计算其在水平角上和竖直角上的角度变化量,由于仪器在旋转上存在惯性、电子元件的温度、电子元件的老化导致水平角、竖直角的量测不准确。因此没有改正的点云数据不能有效地反映地物的真实信息。激光雷达自身的水平角和竖直角的误差,也会影响回波波形。地基激光雷达的大气校正和机载激光雷达气校正也存在着很大的差异性,机载激光雷达的大气校正是需要考虑太阳高度角、方位角,卫星高度角、方位角以及数据采集月份与日期。机载激光雷达的大气校正模型是竖直方向的校正,其中大气中的颗粒大小,颗粒密度都与地基激光雷达大气水平方向上存在很大的差异性。本文针对以上存在的不足,首先解决仪器设备的指向校正问题。具体研究内容包括:仪器设备的水平角、竖直角和测距误差来源,本文设计圆形高反标标靶提取标靶中心点的水平角和竖直角,自主设计特征柱与特征柱载体,采用3D打印技术打印特征柱与特征载体,在特征柱表面粘贴反射片制作成特征点。在仪器设备上布设特征点,通过特征点计算全站仪坐标系到激光雷达站心坐标系的旋转矩阵和平移矩阵,实现全站仪坐标系到仪器站心坐标系之间的转换。利用全站仪获取的圆形标靶的中心值作为基准,根据激光雷达坐标值计算公式,构建误差方程,基于最小二乘原理,计算出仪器设备在水平角和竖直角方向上的系统误差。经过验证改正后的坐标值在Y轴方向上精度提高了1mm,在Z轴方向是精度提高了1mm。然后介绍计算测距改正值的方法,自主设计具有100%反射率的反射板,使用全站仪对反射板的边缘点进行坐标测量,通过特征点计算得到的旋转矩阵和平移矩阵,实现反射板全站仪坐标系到激光雷达坐标系的坐标转换,并且使用转换后的反射板的边缘点坐标构建反射板的平面方程,作为对激光雷达测距校正的约束条件,并将全站仪测距作为基准值,通过激光雷达坐标值计算公式列出误差方程,基于最小二乘原理,计算测距改正值。经过验证改正后的测距值更加的接近全站仪的测距值。
[Abstract]:Lidar has been widely used in the fields of city, traffic, water conservancy and so on, because it can acquire 3D spatial information of object quickly and has the advantages of all-weather, high precision, low cost, etc. Therefore, the accuracy of the calibration also attracted many scholars. The existing ranging correction is based on the correction of waveform parameters, but because of the closure of the original waveform data, the methods of wave peak detection and waveform correction are not clear. The existing waveform parameters extraction is based on Gao Si mixed model to extract the parameters of echo waveform, but this model assumes that there is nonlinearity with waveform data, non-normal distribution form and wave tail uplift. As a result, the precision of waveform parameter extraction is limited, but the relative radiation correction method can not directly obtain the reflection characteristics of ground objects. The method of ranging correction based on waveform parameters does not take into account the attenuation factor of instrument and equipment. In addition, the two-dimensional lidar calculates the angular variation at the horizontal and vertical angles through current changes. Due to the inertia in the rotation of the instrument, the temperature of the electronic components, and the aging of the electronic components, the horizontal angles are caused. The measurement of vertical angles is inaccurate. Therefore, the point cloud data without correction can not reflect the real information of the objects effectively. The errors of the horizontal and vertical angles of the lidar also affect the echo waveform. There is also a great difference between the atmospheric correction of ground-based lidar and that of airborne lidar. The atmospheric correction of airborne lidar needs to consider the solar altitude angle, azimuth angle and satellite altitude angle. Azimuth and data acquisition month and date. The atmospheric correction model of airborne lidar is vertical correction, in which the particle size and particle density in the atmosphere are very different from those in the horizontal direction of the ground-based lidar. In view of the above shortcomings, this paper first solves the problem of pointing correction of instruments and equipments. The specific research contents include: horizontal angle, vertical angle and ranging error source of the instrument and equipment. In this paper, the horizontal angle and vertical angle of the center point of the target are extracted from the circular high inverse target, and the characteristic column and characteristic column carrier are designed independently. Feature points are made by using 3D printing technology to print feature columns and feature carriers and affixing reflectors on the surface of feature columns. In order to realize the transformation from total station coordinate system to instrument center coordinate system, the rotation matrix and translation matrix of total station coordinate system to laser radar station center coordinate system are calculated. Using the center value of the circular target obtained by the total station as the datum, according to the calculation formula of the lidar coordinate value, the error equation is constructed. Based on the least square principle, the systematic errors of the instrument in the horizontal and vertical angles are calculated. The accuracy of the corrected coordinates is improved by 1mm in the Y-axis direction and 1mm in the Z-axis direction. Then it introduces the method of calculating ranging and correcting positive value, designs the reflective plate with 100% reflectivity, uses the total station instrument to measure the coordinate of the edge point of the reflector, and calculates the rotation matrix and the translation matrix through the characteristic point calculation. The coordinate transformation from the total station coordinate system of reflector to the coordinate system of lidar is realized, and the plane equation of the reflector is constructed by using the coordinate of the edge point of the converted reflector as the constraint condition for the ranging correction of lidar. Taking the total station ranging as the reference value, the error equation is listed through the calculation formula of the lidar coordinate value. Based on the least square principle, the ranging correction value is calculated. After verification, the range value is closer to that of total station.
【学位授予单位】：江苏师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P225

【相似文献】