Abstract:Aerial images have features of high resolution, complex background, and usually require large amounts of calculation, however, most algorithms used in matching of aerial images adopt the shallow hand-crafted features expressed as floating-point descriptors (e.g., SIFT (Scale-invariant Feature Transform), SURF (Speeded Up Robust Features)), which may suffer from poor matching speed and are not well represented in the literature. Here, we propose a novel Local Deep Hashing Matching (LDHM) method for matching of aerial images with large size and with lower complexity or fast matching speed. The basic idea of the proposed algorithm is to utilize the deep network model in the local area of the aerial images, and study the local features, as well as the hash function of the images. Firstly, according to the course overlap rate of aerial images, the algorithm extracts the local areas for matching to avoid the processing of redundant information. Secondly, a triplet network structure is proposed to mine the deep features of the patches of the local image, and the learned features are imported to the hash layer, thus obtaining the representation of a binary hash code. Thirdly, the constraints of the positive samples to the absolute distance are added on the basis of the triplet loss, a new objective function is constructed to optimize the parameters of the network and enhance the discriminating capabilities of image patch features. Finally, the obtained deep hash code of each image patch is used for the similarity comparison of the image patches in the Hamming space to complete the matching of aerial images. The proposed LDHM algorithm evaluates the UltraCam-D dataset and a set of actual aerial images, simulation result demonstrates that it may significantly outperform the state-of-the-art algorithm in terms of the efficiency and performance.

Airborne Image Matching Based on Fine-Grained Features and Mutual Attention

Aerial image matching based on dynamic convolution and adaptive neighborhood consensus network

Robust Optical and SAR Image Matching Using Attention-Enhanced Structural Features

A Robust Multisource Remote Sensing Image Matching Method Utilizing Attention and Feature Enhancement Against Noise Interference

Multiview Image Matching of Optical Satellite and UAV Based on a Joint Description Neural Network

Multimodal Remote Sensing Image Matching via Learning Features and Attention Mechanism

Airborne Small Target Detection Method Based on Multimodal and Adaptive Feature Fusion

A Hierarchical Consensus Attention Network for Feature Matching of Remote Sensing Images

Image Matching and Localization Based on Fusion of Handcrafted and Deep Features

Remote Sensing Image Registration Based on Multi-Head Self-Attention Mechanism

Visible-infrared image patch matching based on attention mechanism

A multi-level image alignment method for aerial image and road-based geo-parcel data

Attention-based neural network with Generalized Mean Pooling for cross-view geo-localization between UAV and satellite

Cross-view UAV image matching and localization using deep convolution features

Fine-grained image classification method based on hybrid attention module

An Image Stereo Matching Algorithm with Multi-Spectral Attention Mechanism

UNSUPERVISED MULTI-CONSTRAINT DEEP NEURAL NETWORK FOR DENSE IMAGE MATCHING

Shared contents alignment across multiple granularities for robust SAR-optical image matching

Attention-Based Monocular Depth Estimation Considering Global and Local Information in Remote Sensing Images

Multimodal Remote Sensing Image Matching Combining Learning Features and Delaunay Triangulation

Local Deep Hashing Matching of Aerial Images Based on Relative Distance and Absolute Distance Constraints.