Image retrieval method based on hierarchical convolutional neural network

Abstract

The invention discloses an image retrieval method based on a hierarchical convolutional neural network, and mainly aims at solving the problem that in existing all-sky aurora image retrieval, the accurate rate is low. The method comprises the implementation steps that 1, local key points of all-sky aurora images are determined by adopting an adaptive polar barrier method; 2, local SIFT features ofthe all-sky aurora images are extracted, and a visual vocabulary is constructed; 3, the convolutional neural network is pre-trained and subjected to fine tuning, and a polar region pooling layer is constructed; 4, region CNN features and global CNN features of the all-sky aurora images are extracted; 5, all the features are subjected to binarization processing, and hierarchical features are constructed; 6, a reverse index table is constructed, and the global CNN features are saved separately; and 7, hierarchical features of a queried image are extracted, the similarity between the queried image and the database images is calculated, and a retrieval result is output. According to the method, matching of the local key points is achieved through the hierarchical features, the problem that inan existing image retrieval method, the false alarm rate is high is solved, the advantage of being high in retrieval accuracy rate is achieved, and the method is suitable for real-time image retrieval.

Description

technical field [0001] The invention belongs to the technical field of image processing, relates to a deep learning algorithm and image retrieval technology, and can be used for accurate retrieval of large-scale aurora images. Background technique [0002] The aurora is a high-latitude natural luminous phenomenon produced by high-energy charged particles carried by the solar wind that settle along the geomagnetic field lines and collide with the particles in the earth's atmosphere. Therefore, establishing an efficient image retrieval system to complete the screening of effective data and the analysis of key data in large-scale auroral images can help humans obtain a large amount of information about solar-terrestrial space activities. [0003] Since the aurora has significant research value to the solar-terrestrial space, humans have detected it through various means in recent years. Among them, ground-based optical imaging detection is an important project of polar scienti...

AI Technical Summary

Problems solved by technology

[0006] However, the existing aurora image retrieval methods still use artificially designed feature extraction methods, which lack data learning process and semantic information cognition, resulting in poor feature recognition; and a single global or local analysis mode makes the retrieval results easy to ignore details or fall into the trap. Local optimization; at the same time, the increasingly large database will inevitably require further optimization of the query index structure to ensure low memory consumption and achieve accurate retrieval of large-scale aurora images

Images