The SVM classification leafminer-infected leaves based on fractal dimension

Leafminer is one of pest of many vegetables, and the ordinary way to control it is spraying the pesticides. Because of polluting the environment and remaining in the vegetable, the pesticides are restricted to use. In order to get the information of the pest in the vegetable before the damage was not serious, this research used the image processing technology and the fractal dimension to work out the damaged degrees of the leafminer- infected cucumber leaves. Two different kernel functions have been used to set up the classifying models, and the fractal dimension of the damaged cucumber leaf images was applied to the Threshold methods and the SVM neural network performing the recognition and classification. The numerical experimental results showed that the classification precision of the Threshold methods is the best (61%) at the (3,7) computing step, and is the worst ( 36%) at the (3,6) computing step, and the error number of the Threshold methods is very high (the worst is 18 times at the (3,6) computing step). The precisions of the SVM are beyond 80% by using the polynomial-based kernel function and 90% via using the RBF kernel function, and the RBF kernel based SVM excels to the polynomial-based kernel. The SVM-based fractal dimension analysis models can be used to classify the leafminer-infected cucumber leaves.