Grey cluster estimating model of soil organic matter content based on hyperspectral data

As to the uncertainty relations between soil organic matter content and spectral characteristics, at first, based on the objective function that the sum of squares of generalized weighted grey distance is minimum, this paper proposes a new self-iteration grey clustering model whose classification standard is unknown. It then establishes a grey clustering estimating model of soil organic matter content based on hyperspectral data, and then applies the model to Hengshan County of Shanxi Province. The results show that the self-iteration grey clustering model can not only make full use of the intrinsic information of clustering object indicators but also utilize expert knowledge and experience, and overcome the subjectivity of determining classification standards and weights. The average whitening and grey prediction accuracy of test samples is 93.088% and 99.192% respectively. The example shows that the presented model is valid.