机械耕作的土壤结构数字图像分析

发布时间：2018-04-04 22:03

本文选题：土壤结构　切入点：机械耕作　出处：《南京农业大学》2015年硕士论文

【摘要】：土壤耕作的目的是创建良好的苗床,而对土壤结构或苗床质量的精确定量又是正确评价耕作机具作业性能的重要途径。目前评价苗床质量多用感官评定,如使用现场打分评价或土壤视觉打分等感官评价农业土壤的质量状况,但是感官评价仍存在主观性较强、规范性差等状况。用于评价耕作机具的碎土性能及土壤耕作质量的常规指标是碎土率及土壤破碎体尺度分布等指标,通常较少对耕作生成的土壤结构进行精细分析,然而精细的土壤结构分析更利于提供丰富的信息。本文使用数字图像分析技术定量土壤结构并提供3个外观几何指标,进而分别从土壤结构的棱角性、形状指数、矩形度、尺度分布4个方面评价土壤结构。本研究首先探讨土壤结构的数字图像处理方法与指标,针对自然地理过程形成的土壤结构,研究其图像获取及计算分析方法。对土壤结构的外观几何指标(棱角性、形状指数、矩形度)进行计算,定量了自然地理过程形成的土壤结构特征。结果表明自然地理过程作用下黄棕壤呈块状分布,且随土壤结构的尺度增大其形状指数及棱角性呈增大的趋势,而矩形度则不断降低。将提供的数字图像处理方法分别应用于水稻土的犁耕和旋耕生成的土壤结构分析。对比冬夏季犁耕和旋耕的土壤结构,统计分析棱角性、形状指数以及矩形度指标。结果表明棱角性和形状指数能很好地区分不同耕作方式的土壤结构,而矩形度这个指标不能很好地区分耕后土壤结构。对土壤结构的图像信息与体积(质量)信息的关系进行进一步研究,将不同尺度区间的土壤结构体分别进行不同俯视角度拍摄,建立投影面积与质量的对应关系,然后在此基础上分析土壤结构体的质量-尺度分布。试验结果表明在俯视角60°条件下,各尺度区间的土壤结构体投影面积与质量拟合效果都是最好。用数字图像法分析水稻土犁耕耕作下土壤结构质量-尺度分布模型为F(dD)=(d/252.102)0.77。最后再进一步探究图像法原位测试耕作的土壤结构,研究图像获取及计算方法,并将原位测试获取的土壤结构指标与非原位的进行对比。结果表明两种方法各几何指标存在很好的相关性,但是各指标也存在差异性,尤其是矩形度的差异显著。本文所提供的数字图像分析方法快捷方便,能够提供更多的土壤结构信息,所得土壤结构尺度分布模型更加准确。
[Abstract]:The purpose of soil tillage is to create a good seedling bed, and the accurate quantification of soil structure or seedling bed quality is an important way to correctly evaluate the working performance of tillage machines and tools.At present, sensory evaluation is used to evaluate the quality of seedling bed, such as field scoring or soil visual scoring, etc., but sensory evaluation still has strong subjectivity and poor standardization.The conventional indexes used to evaluate the soil fragmentation performance and soil tillage quality of tillage tools are the rate of soil fragmentation and the size distribution of soil fragments, and the fine analysis of soil structure produced by tillage is usually rare.However, fine soil structure analysis is more helpful to provide rich information.In this paper, the digital image analysis technique is used to quantify soil structure and provide three geometrical indexes of appearance, and then the soil structure is evaluated from four aspects of the soil structure, such as prism, shape index, rectangle degree and scale distribution.In this study, the digital image processing method and index of soil structure were discussed, and the methods of image acquisition and analysis were studied for the soil structure formed by natural geographical process.The appearance geometric indexes of soil structure (angularity, shape index, rectangle degree) were calculated, and the soil structure characteristics formed by natural geographical process were quantified.The results show that yellow brown soil distributes in blocks under the action of natural geography, and its shape index and angularity increase with the scale of soil structure increasing, but the rectangle degree decreases continuously.The digital image processing method is applied to the soil structure analysis of paddy soil produced by ploughing and rotary tillage, respectively.Comparing the soil structure of ploughing and rotary tillage in winter and summer, the angularity, shape index and rectangle index were analyzed statistically.The results showed that the prism and shape index could distinguish the soil structure of different tillage methods well, but the rectangular degree index could not distinguish the soil structure after tillage very well.The relationship between the image information of soil structure and the information of volume (quality) is further studied. The corresponding relation between projection area and mass is established by taking different overlooking angles of soil structure with different scales.Then the mass-scale distribution of soil structure is analyzed.The experimental results show that the fitting effect of the projection area and the quality of soil structure is the best under the condition of the overlooking angle of 60 掳.The numerical image method was used to analyze the soil structure quality and scale distribution model under ploughing tillage in paddy soil.Finally, the soil structure of in situ testing tillage by image method is further explored, the method of image acquisition and calculation is studied, and the index of soil structure obtained by in situ test is compared with that obtained by in situ test.The results show that there is a good correlation between the two methods, but there are also differences among them, especially the difference of rectangular degree.The digital image analysis method provided in this paper is quick and convenient, and can provide more information of soil structure, and the scale distribution model of soil structure is more accurate.
【学位授予单位】：南京农业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：S152.4

【相似文献】