A Line Simplification Method Based on Support Vector Machine

Line simplification is one of the most important parts of automatic map generalization. Aiming at the problem that the effect of the current line simplification methods is too dependent on the artificial setting of algorithm parameters and thresholds, and lacking the adaptive learning ability for different simplification environments, this paper considers the line simplification as a binary classification problem about the selection or deletion of the simplification unit. From the perspective of case-based studying, a new line simplification method based on support vector machine (SVM) is proposed. Firstly, point and bend are taken as the simplification unit, the simplification case is automatically obtained from the simplification results of experts, and then the feature description items are extracted as the attribute space of the case sample. The line simplification classifier is trained by SVM. Finally, the SVM classifier is used to classify the simplification units of the new line elements into selection and deletion, so that the line simplification is realized. The experimental results show that this method can well reduce the simplification intention of the experts by learning the simplification case, with a high classification accuracy for the selection and deletion of simplification units, and can adaptively complete the line simplification.