Image retrieval method and device thereof

Abstract

The invention discloses an image retrieval method and a device thereof, and the method comprises the steps: inputting a to-be-retrieved image into a pre-trained deep learning classification network, extracting N-dimensional CNN features from the to-be-retrieved image, recalling a first preset number of candidate similar images from a retrieved image set based on the N-dimensional CNN features, andthen extracting local features from the to-be-retrieved image; and converting the local features into M-dimensional local features through a bag-of-words model, and recalling a second preset number of target similar images from the candidate similar images based on the M-dimensional local features. Because the N-dimensional CNN features can represent the overall feature description of the to-be-retrieved image, and the M-dimensional local features can represent the local feature description of the to-be-retrieved image, the method achieves the simultaneous combination of the overall feature description and the local feature description of the to-be-retrieved image for image retrieval, and improves the retrieval precision of the image.

Description

technical field [0001] The present invention relates to the technical field of image retrieval, and more specifically, to an image retrieval method and device. Background technique [0002] At present, a huge amount of multimedia data will be generated in the process of using the Internet to communicate, which has caused considerable obstacles for Internet users to quickly find the information they need. Based on this, search technology has emerged as the times require. Facing the ever-emerging demands, how to enable Internet users to efficiently and quickly search for target images among massive images has become an important research direction. [0003] In the prior art, two methods are mainly used for image retrieval. One is to perform image retrieval based on the description of the overall image features. Therefore, when the image partially contains retrieval targets, it is difficult to accurately describe the local features of the image. The other is image retrieval ba...

AI Technical Summary

Problems solved by technology

[0003] In the existing technology, two methods are mainly used for image retrieval. One is to perform image retrieval based on the overall feature description of the image. Therefore, when the image partially contains the retrieval target, it is difficult to accurately describe the local features of the image.

The other is image retrieval based on image local feature description. Although this method can accurately describe the local features of the image, the overall description of the image is lost, resulting in low image retrieval accuracy.

Images