Soil Roughness Measuring System Combined With Image Processing

Abstract

The soil surface roughness is measured by a self-made soil roughness meter, and the image obtained is digitally processed by using image processing software MATLAB R2014a to obtain two main parameters for describing the soil roughness, that is, the root mean square height S and correlation length L, in which the higher the root mean square height, the smaller the correlation length, the greater the roughness of the soil surface. Experimental results show that this method is feasible and the change trend of the root mean square height at each sampling point on the whole is opposite to that of the correlation length. Simultaneously, the experimental method reduces the requirements on the environment, including wind and sunlight, and the achieved accuracy can be controlled within 0.5 cm. Finally, compared with the manually read soil height and the calculated root mean square height and correlation length, the root mean square height measurement error at each sample point is within 0.5 cm, and the total error percentage is within 2.03%; while the correlation length measurement error is within 0.4 cm, and the total error percentage is 0.22%. They can satisfy the real-time measurement requirements of soil roughness.