The rapid development of unmanned aerial vehicle (UAV) aerial survey technology and deep learning has revolutionized traditional aerial survey methods. With its fast measurement, low cost, good safety, and high reliability, UAVs can quickly obtain image data of a certain survey area in a relatively short time. Although UAV surveying has been used in various fields of related industries, there are still certain limitations to quickly obtaining desired information from a group of images. Therefore, this paper combines the VGG16 deep learning model to match UAV images and make it applicable to the web environment. Starting with extracting image features of UAVs from the VGG16 model, this paper implements a UAV image feature extraction and retrieval system based on VGG16. Then, the matching efficiency is compared with the matching model based on the SIFT algorithm. Finally, the image feature extraction and retrieval system based on VGG16 is combined with the web system to study the retrieval of image data of interested areas or image data without spatial information, as well as spatial information queries under web conditions.
Rapid retrieval system of UAV image based on deep learning
Second International Conference on Geographic Information and Remote Sensing Technology (GIRST 2023) ; 2023 ; Qingdao, China
Proc. SPIE ; 12797
2023-08-30
Conference paper
Electronic Resource
English
Learning semantics in content based image retrieval
IEEE | 2003
|End-to-End Learning of Deep Visual Representations for Image Retrieval
British Library Online Contents | 2017
|Sketch Based Image Retrieval using Transfer Learning
IEEE | 2019
|Learning Semantics in Content Based Image Retrieval
British Library Conference Proceedings | 2003
|Deep Semantic Coding for Wireless Image Retrieval
IEEE | 2022
|