Comput. Electron. Agric. | 2019
A new method to analyse the soil movement during tillage operations using a novel digital image processing algorithm
Abstract
Abstract Tillage operations are a vital part of agricultural crop production. Economic and environmental considerations are forcing farmers to manage soil tillage with optimum tool configurations to get the desirable soil condition. Soil disturbance caused by soil engaging tools, in particular soil layer mixing, is an important phenomenon that needs to be clearly understood. Current methods used to investigate soil layer mixing are limited, especially when multiple layer mixing needs to be investigated. The use of physical tracers is the most common approach. Although this method can provide some useful information, since it does not provide a full representation of how soil layers are moving or mixing, error-prone rough estimation is unavoidable. In this study, the mixing of soil layers was investigated using different coloured sands placed in layers at different depths (to continuously investigate the soil layer mixing). A new colour clustering algorithm to analyse such multi-coloured layer mixing performance was developed using K-means clustering with a PCA (principal component analysis) approach. The validation study of the proposed algorithm was conducted using the Columbia multispectral image database before implementation on the experimental coloured soil test. The test results showed that the proposed method and algorithm is effective in analysing multiple soil layer mixing.