Image retrieval method, device, server and storage medium
A technology of pictures and picture groups, applied in the Internet field, can solve problems such as inability to recall pictures and inability to meet query requirements.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0030] figure 1 It is a flow chart of the picture retrieval method provided by Embodiment 1 of the present invention. This embodiment is applicable to the situation of picture retrieval, especially the situation of using a long query or a query with multiple limited words for picture retrieval. The method can be searched by pictures device, which may be implemented in software and / or hardware, and may be configured in a server. Such as figure 1 As shown, the method specifically includes:
[0031] S110. Identify multiple picture groups with the same content from pictures on all web pages.
[0032] With the continuous development of network technology, pictures, an important form of information representation, inevitably appear on various web pages, and the scale of network data continues to expand at any time, and it often appears that multiple different web pages contain one or more content at the same time. The same picture, where the picture with the same content includes...
Embodiment 2
[0043] figure 2 It is a flow chart of the image retrieval method provided by Embodiment 2 of the present invention. This embodiment is further optimized on the basis of the foregoing embodiments. Such as figure 2 As shown, the method includes:
[0044] S210. Identify multiple picture groups with the same content from pictures on all web pages.
[0045] S220. Filter and deduplicate all source webpages of each picture in each picture group, and aggregate texts related to the pictures of the remaining source webpages to obtain a text description of each picture group.
[0046] The same image will appear on multiple different web pages. If the duplication is not removed, the indexed content will be redundant. At the same time, the surrounding text of some pictures from low-quality pages is not related to the picture, and if not processed, it will also affect the recall quality. Preferably, all source webpages are screened and deduplicated, the source webpage with the highest...
Embodiment 3
[0061] image 3 It is a flow chart of the image retrieval method provided by Embodiment 3 of the present invention. This embodiment is further optimized on the basis of the above embodiments. Such as image 3 As shown, the method includes:
[0062] S310. Identify multiple picture groups with the same content from pictures on all web pages.
[0063] S320. Filter and de-duplicate all source webpages of each picture in each picture group, aggregate the picture-related texts of the remaining source webpages, and obtain a text description of each picture group.
[0064] S330. Based on the text description of each picture group, establish an inverted index for each picture in each picture group, wherein, for each text description, the inverted index includes at least all The webpage from which the text description corresponds.
[0065] S340. Obtain the input search term.
[0066] S350. Recall at least one picture according to the correlation between the search term and the text...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com