关键词: 土壤, 断层成像, 孔隙检测, 孔隙结构, 孔隙数量

Abstract: Aiming at the problems of low precision, poor automation and high detection cost in the current soil or matrix internal pore structure detection system, an automatic detection system with low-cost soil pore structure is designed. The system's structure comprises the automatic grinding module, cross-section image acquisition module, and motion control module, which can continuously and automatically grind the soil or substrate samples with high-precision cross-sections and obtain images with sequence information. The pore distribution information can be extracted based on the cross-section image, which is processed by cutting, enhancing, and segmenting. The test results show that the automated detection system designed in this paper automatically calculates the number of macropores and effective pores on each tomography image by an optimized image processing method. The detection system outperforms CT scanning in terms of grinding cross-section thickness and image resolution, which can reach 0.05 mm and 24 million pixels respectively, and has a low usage cost. In addition, the system uses optimized image processing method for cross-sections. It can automatically count the number of macropores and effective pores on each cross-section based on the images. The automatic detection system designed in this paper provides a new technical means and equipment support for soil pore research.

Key words: soil, tomography, pore detection, pore structure, pores quantity