We applies Hu invariant moments and edge direction histogram in the binary trademark images to extract the shapes of objects. Then, a similarity measure called grade measure is provided to conduct the matching process based on the rank of the feature distances. Finally, a feedback algorithm to determine the weight of each feature according to the users' response was presented. From the experiences, it is obvious that this method is consistent with human's vision semanteme, and do well in trademark images retrieval.
參考文獻(xiàn)
相似文獻(xiàn)
引證文獻(xiàn)
引用本文
盧章平,朱科鈐.基于形狀特征和用戶反饋的商標(biāo)圖像檢索技術(shù)[J].農(nóng)業(yè)機(jī)械學(xué)報(bào),2007,38(5):150-152.[J]. Transactions of the Chinese Society for Agricultural Machinery,2007,38(5):150-152.