Design And Implementation Of Upper Computer Software For Wireless Sensor Networks Localization System

The macroscopic physical-mechanical properties and microscopic characteristic of undisturbed soil is related closely. The analysis of key elements of soil microstructure is critical in the analysis of macroscopic physical-mechanical properties. It is of great significance and widely applied value in the analysis of microscopic characteristic of soil.The loess electron microscopy(SEM) image is chosen as the study object to realize the extraction and analysis of loess microstructure. The key elements including pore, silt and clay of loess microstructure are extracted based on the digital image processing technology, and simple calculation and quantitative analysis are conducted to make feature analysis of key elements of loess microstructure, which provides assistance of the analysis of loess macroscopic physical-mechanical properties. The main content of the study includes three parts as follows:(1)The SEM images pre-processing of key elements of loess microstructure are conducted, which includes image segmentation and scale length extraction. The influence of the key elements of the loess microstructure that caused by scale region can be reduced by image segmentation, and the edge detection and Hough line detection are used in scale length extraction.(2)The key elements of loess microstructure are extracted, including pore, silt and clay and so on. The methods such as pre-processing denoising, histogram statistic and peak threshold segmentation are applied on the micro image of loess in the processing of extracting the pore. The opening pre-processing, histogram statistic and peak threshold segmentation are used in the extracting of silt. The clay group is extracted based on the results of pore and silt extraction.(3)The contour extraction and calculation of pore and silt extraction results are conducted. The opening algorithm is used to eliminate the isolated points of pore extraction, and the contour extraction is applied on the opening results. The contour extraction of silt extraction is used directly for there is no isolated points existed. The simple quantitative analysis of pore and silt extraction results is proceed and compared with the artificial hand-pointed tracing.By comparing the experiment result with the hand-painting images, the area ratio error of pore and silt is in the range of 3%- 11%. A further analysis of pore and silt images is accomplished by IPP software to get the quantitative microstructural parameters of different undisturbed loess and the composition features of it is analyzed. The experiment results show that the largest average diameter of Lanzhou loess silt can reach to 22.25 ?m, and the shape of it is most deviated from a circular. The particles of three regions are spatial distributed uniformly. The average pore diameter of loess in Lanzhou is closed to Xi’an with the average diameter of 19.94 ?m and 18.58 ?m, and more deviated from roundness. The average diameter of silt is close to the average pore diameter, and the shape of silt is closer to a circle.