针对庞大而复杂的外观专利数据给外观专利设计带来的审查问题,对基于内容的图像检索、自动分类和自学习等关键技术进行了分析,同时对图像检索系统的性能评价指标体系进行了阐述。
With more and more acceptance and authorization of design patents applications in China.it is a great challenge to process large and complex patent data on patent examination of appearance design is analysed.This paper discusses and analyzes key technology on design patents examination such as content-based image retrieval,automatic classification,machine learning and so on.A complete performance index of image retrieval system is also described in detail in this paper.