Research on Content-based Commodity Image Retrieval Technology in E-commerce
Jiang Da-yang · Xinxi wangluo anquan · 2013
Currently, the deficiency of text-based image retrieval technology (TBIR) is becoming more and more obvious in the rapid development of e-commerce. Content-based image retrieval technology (CBIR) is the driving force of the reform of the future e-commerce retrieval technology due to its application value and realistic significance in business. The combined image feature retrieval technology based on one-dimensional color feature vector and shape feature vector overcomes the deficiency of the previous single feature vector retrieval technology. The combined application of two feature vectors may realize the matching search results with higher similarity, so as to meet the practical retrieval demands of CBIR in e-commerce.