Title:
REMOTE SENSING IMAGE OVERALL PLANNING RECOMMENDATION METHOD BASED ON CONTENT UNDERSTANDING
Document Type and Number:
WIPO Patent Application WO/2024/037585
Kind Code:
A1
Abstract:
Disclosed in the present invention is a remote sensing image overall planning recommendation method based on content understanding. The method comprises the following steps: S1, acquiring a query condition and a first image set; S2, acquiring depth features of the first image set; S3, acquiring local features of the first image set; S4, acquiring fusion features of the first image set; S5, acquiring fusion feature similarities among images in the first image set; S6, selecting an initial image from the first image set, and determining optimal indexes, the optimal indexes comprising a feature similarity and an effective coverage rate; S7, using the initial image as a root node, performing split expansion for multiple times from the root node according to the optimal indexes to obtain a complete split search tree, and obtaining a second image set on the basis of the split search tree; and S8, recommending the second image set to a user. According to the method, the fusion feature similarities of image sets and the effective coverage rates of images are used as optimal indexes to perform split tree search, such that an image set to be recommended can be rapidly acquired.
Inventors:
CHEN YU (CN)
PENG ZHE (CN)
ZOU SHENGBING (CN)
PENG ZHE (CN)
ZOU SHENGBING (CN)
Application Number:
PCT/CN2023/113515
Publication Date:
February 22, 2024
Filing Date:
August 17, 2023
Export Citation:
Assignee:
BEIJING DATA INTELLIGENCE INFORMATION TECH CO LTD (CN)
International Classes:
G06F16/583; G06T7/11
Foreign References:
| CN115248876A | 2022-10-28 | |||
| CN110717058A | 2020-01-21 | |||
| CN106326288A | 2017-01-11 | |||
| CN114078137A | 2022-02-22 | |||
| CN113378842A | 2021-09-10 | |||
| US20180218429A1 | 2018-08-02 |
Download PDF: