Environmental Art Design Entity Matching Method Based on the Outlier Detection

For the issue of the environmental art design entity matching with the diversity of expressions, usually there is no concise and effective matching rule existing, and it is also very difficult to train the precise classification model. Aiming at this issue, this paper has proposed an automatic environmental art design entity matching method based on the Outlier Detection, which is recorded as DMacth algorithm. Firstly calculate recording ordered pair similarity on the matching property, and map the ordered pair as the points in the feature space; then estimate the outlier distance of each ordered pair in the feature space; finally, according to the outlier distance and the matching constraint, extract the matching ordered pair. The experiment results have verified the validity of the DMacth method.