Transfer learning-based multi-view commodity image retrieval and identification method

Abstract

The invention discloses a transfer learning-based multi-view commodity image retrieval and identification method. The method comprises the steps of 1, establishing a multi-view image basic library according to a commodity list, performing fine adjustment on a pre-trained deep residual error network by using a small amount of commodity images through a transfer learning technology, extracting features of the image basic library by using the network, performing dimension reduction on the features, constructing a feature library, and finally according to corresponding relationships among the feature library, the image basic library and commodity types, establishing a mapping table; 2, after to-be-identified commodity images are obtained, extracting features of the images by using the networkand performing dimension reduction; and 3, performing distance measurement on the features of the to-be-identified commodity images and the features of the images in the basic library, taking the mostsimilar image with the shortest distance as a matching result, and through the mapping table, obtaining commodity type names of the to-be-identified commodity images. The features with strong representation capabilities can be automatically extracted; a semantic gap is further broken through; and the retrieval efficiency and the identification precision are improved by only utilizing a small amount of image basic libraries and low-dimensional features.

Description

Technical field: [0001] The invention belongs to the application field of image processing, retrieval and recognition, and specifically proposes a multi-view commodity image retrieval and recognition method based on transfer learning. Background technique: [0002] In the era of Web 3.0, especially with the popularity of social networking sites and shopping sites, heterogeneous data such as images, videos, and audios are increasing day by day. For example, the number of pictures stored on the back-end system of Taobao, the largest e-commerce system in my country, has exceeded 28.6 billion. Therefore, how to conveniently, quickly and accurately retrieve and identify the images that users are interested in in the vast image database containing rich visual information has become a research hotspot in the field of multimedia information. For this reason, content-based image retrieval technology has been gradually established and developed rapidly in recent years, and has been w...

AI Technical Summary

Problems solved by technology

However, the existing content-based image retrieval also has some defects, mainly as follows: (1) There is a semantic gap between feature description and high-level semantics that is difficult to fill, although deep convolutional neural network (CNN, Convolutional Neural Network) The dominant feature expression method has also begun to be developed on the image retrieval of the same object, but it is still necessary to introduce a network that can extract features with stronger expressive capabilities, so as to further break through the semantic gap; (2) Massive image libraries and their corresponding high-dimensional The storage of the feature database occupies a large amount of storage space; (3) due to the large-scale image database and high-dimensional feature database, it is difficult to meet the real-time requirements of the system by directly using similarity measurement and indexing strategies; (4) Some image retrieval systems, such as Alibaba’s Pailitao, only have the retrieval function, but do not have the function of identifying product categories

Images