Digital Image Analysis Of Tillage-induced Soil Structure

The primary objective of soil tillage is to prepare a good seedbed.A precise quantification of soil structures or seedbed quality is a critical basis for a correct evaluation of the performance of tillage implements.In most cases,seedbed quality was visually assessed,e.g.agricultural soil was graded in situ or evaluated with visual assessment.These evaluation methods are subject to be biased and lack of uniformity due to subjective errors,which could be introduced by individual observers.The most commonly used parameters for soil fragmentation performance of tillage tools include soil fragmentation ratio and soil fragment size distribution.Less attention has been paid on the detailed analysis of soil structures.When soil structures were assessed with refined resolutions,more structure-related information can be derived.In this research digital image processing was performed to evaluate soil structures and 3 geometric parameters for soil structure quantification were supplied.This allows soil structures to be quantified with 4 indices,i.e.angularity,shape index,rectangularity and fragment size distribution,respectively.Suitable method and parameters of digital image processing for soil structure analysis were investigated,which,including the methodology of image collection and calculation,were used for assessing the soil structures resulted from natural pedological processes.Geometrical parameters,e.g.angularity,shape index and rectangularity of soil structures,were calculated to quantify these naturally evolved soil structures.The yellow-brown soil under natural pedological process is blocky structured and,as the size increases,both shape factor and angularity increases.Whereas rectangularity decreases steadily.The proposed method of digital image processing was subsequently applied to structure analysis on paddy soil,where both the ploughing-induced and rotary-tilled soil structures were compared.Soil structures collected in both winter and summer season were compared.The angularity,shape factor and rectangularity were statistically evaluated,revealing the fact that angularity and shape factor were 2 parameters most suitable for soil structure analyzing,while the rectangularity was poor in evaluating tillage-induced soil structures.A further investigation on the relationship between image and volume/mass data of soil structures was made by photographing the soil structures with different depression angles.The projection areas of soil structures were correlated with respective masses.This allowed the mass-scale of soil structures to be quantified.It was noted that,under 60°depression angle,the best fit was observed between soil structure projection areas and respective masses,regardless of the size ranges of soil structures.The acquired model of mass-scale distribution of tillage-induced soil structure with ploughing in paddy soil satisfies:F(d<D)=(?)Soil structure was analyzed in situ with the proposed technique of image processing.The result was compared with soil structure sampling method.The result showed that,for both methods,good correlation exists among geometric indices.But differences also exits,especially for rectangularity,which was proved significant.The proposed method of digital image processing for soil structure analysis was quick,simple and efficient,providing more detailed information of soil structures.With this tool for soil structure analysis,the acquired model of soil structure size distribution could be more precise.