Content-based shop identification image retrieval method

Abstract

The invention discloses a content-based shop identification image retrieval method. The method comprises the following specific steps of: 1, acquiring an image; 2, image detection: identifying a shopidentification image from the acquired images; 3, preprocessing the shop identification image and the database image data; 4, extracting features of the shop identification image and the database image; and 5, carrying out similarity comparison on the shop identification image and the database image, and if the similarity is higher than a set threshold value, determining that the image is a retrieved image.

Description

technical field [0001] The invention relates to the field of computing retrieval computing, in particular to a content-based shop logo image retrieval method. Background technique [0002] In today's society where information technology is so developed, all kinds of information and data are flooding everyone's eyes, such as multimedia information such as video images; the amount of information is increasing, and it is becoming more and more difficult to retrieve effective data from large-scale data. The larger the image size, the more various retrieval methods have appeared; the traditional image retrieval technology based on text description needs to manually mark the image content with the keywords specified by the image, and search according to the keywords; however, everyone's understanding of the image content are different, which will cause keyword matching errors and reduce the accuracy of image retrieval; content-based image retrieval is to find the retrieval image w...

AI Technical Summary

Problems solved by technology

[0002] In today's society where information technology is so developed, all kinds of information and data are flooding everyone's eyes, such as multimedia information such as video images; the amount of information is increasing, and it is becoming more and more difficult to retrieve effective data from large-scale data. The larger the image size, the more various retrieval methods have appeared; the traditional image retrieval technology based on text description needs to manually mark the image content with the keywords specified by the image, and search according to the keywords; however, everyone's understanding of the image content are different, which will cause keyword matching errors and reduce the accuracy of image retrieval; content-based image retrieval is to find the retrieval image with the highest similarity in the database by inputting the image to be queried as input data; The registered store logo image is a high-definition full image, while the image to be queried is an image taken with a mobile phone or VR glasses, and the logo only occupies a part of the image. There are image quality problems, shooting angles, changes in light and background, so It is difficult to accurately retrieve whether the identifier to be retrieved has been registered in the database

Images