土壤表面粗糙度是一项重要的土壤物理参数，已有的各种测量方法存在测量效率和误差难以兼顾的问题。为了在一定精度下提高野外测量工作效率，该研究基于红外结构光技术设计了一套便携的土壤表面粗糙度测量系统。该系统主要包括红外结构光扫描仪、便携式计算机、支架等，具有3.2 mm空间分辨率和3 mm的距离分辨率的性能。通过水平面板测量试验，发现本系统相对误差较小，相对误差最小区域中0.5个测量单位以内的误差（e≤1.5 mm）占87.87%，1个测量单位以内的误差（e≤3 mm）占99.58%，而大于1个测量单位的误差仅占0.42%。通过土壤表面粗糙度测量试验，发现本系统绝对误差较明显，其测量结果低于1 mm 分辨率的土壤粗糙度值。通过误差分析发现：该系统的测量误差包括固有误差和随机噪声，呈特定的规律性分布；导致绝对测量误差的主要原因为该系统的性能；由于系统的随机噪声，基于水平面板距离图像的土壤表面距离图像校正存在不确定性。该研究结果为进一步降低红外结构光三维技术测量土壤表面粗糙度的误差提供了依据。%Soil surface roughness (SSR) is a very important parameter for describing soil physical characteristics. It is widely used in wind and water erosion studies, and for retrieving soil moisture using passive or active microwave sensing data. There are several methods to quantify SSR. The techniques of quantifying soil surface roughness mainly include the pin meter and profile meter, photography, and laser scanning. However, these methods respectively have some limitations. The main disadvantage of the pin meter and profile meter is the potentially destructive effect while using these methods. Laser scanning equipment is expensive, bulky, and therefore, it is hard to work in field conditions. Photography only acquires 2D information not real 3D information of the soil surface. In recent years, structured light measurement and range image technology has been developed very well, and the main equipment of this technology became cheaper and more portable. It has been widely used in many aspects for 3D surface reconstruction. In this study, a system based on an infrared structured light 3-dimensional technique was designed for a more portable and efficient measuring of SSR. The system mainly contains an infrared structured light scanner, a tripod with a beam, a portable computer, and a level board. The scanner contains two main parts, a structured light projector and a sensor which receives the reflected structured light. The computer with installed special software was used to control the scanner and store the data. A measurement experiment was conducted. A plain board and two different soil surfaces were measured. In a practical measuring, the scanner was fixed on one end of the beam of a tripod, and directed the front of scanner towards the soil surface, and was about 100cm high from the soil surface. The spatial resolution depended on the distance between the object and the scanner. A range image of a level board was treated as the horizontal reference for correct those range images of the soil surface. With a practical measuring process, it was found that the portability and capability of acquiring data of a structured light 3D sensor was excellent. But the measuring results had some errors because of the limitation of the spatial resolution of the system. The precision of the system needs to improve. From the analysis of the errors of the plain board measuring, we can draw some conclusions as follow. (1) The errors of the range images and surface information included inherent error and random noise, and had a specific distribution. The errors in the middle of the image were lower than in the other parts. (2) Compared with the accurate results with a higher resolution, the soil surface elevation and the roughness that was acquired by this system had less precision, and the precision was determined by the performance of the structured light scanner. (3) Due to the random noise, the correction of the soil surface elevation using the measurements of the plain board had uncertainty. The performance of this system was tested by a practical measuring experiment. It is believed that this method would be popular in SSR measuring after doing some improvements in next investigation according some findings in this study.
展开▼
