Abstract:Co-registration of aerial imagery and Light Detection and Ranging (LiDAR) data is quite challenging because the different imaging mechanisms produce significant geometric and radiometric distortions between the two multimodal data sources. To address this problem, we propose a robust and effective coarse-to-fine registration method that is conducted in two stages utilizing spatial constraints and Gabor structural features. In the first stage, the LiDAR point cloud data is transformed into an intensity map that is used as the reference image. Then, coarse registration is completed by designing a partition-based Features from Accelerated Segment Test (FAST) operator to extract the uniformly distributed interest points in the aerial images and thereafter performing a local geometric correction based on the collinearity equations using the exterior orientation parameters (EoPs). The coarse registration aims to provide a reliable spatial geometry relationship for the subsequent fine registration and is designed to eliminate rotation and scale changes, as well as making only a few translation differences exist between the images. In the second stage, a novel feature descriptor called multi-Scale and multi-Directional Features of odd Gabor (SDFG) is first built to capture the multi-scale and multi-directional structural properties of the images. Then, the three-dimensional (3D) phase correlation (PC) of the SDFG descriptor is established to detect the control points (CPs) between the aerial and LiDAR intensity image in the frequency domain, where the image matching is accelerated by the 3D Fast Fourier Transform (FFT) technique. Finally, the obtained CPs not only are employed to refine the EoPs, but also are used to achieve the fine registration of the aerial images and LiDAR data. We conduct experiments to verify the robustness of the proposed registration method using three sets of aerial images and LiDAR data with different scene coverage. Experimental results show that the proposed method is robust to geometric distortions and radiometric changes. Moreover, it achieves the registration accuracy of less than 2 pixels for all cases, which outperforms the current four state-of-the-art methods, demonstrating its superior registration performance.

Dual-band Low Level Light Image Real-time Registration Based on Pixel Spatial Correlation Degree

Study on Real-Time Registration in Dual Spectrum Low Level Light Night Vision Technique

Dual-Focal Camera Hdr Imaging Based On Convolutional Neural Network

Parallel tracking and matching for natural feature based AR registration

Infrared and Visible Image Registration in UAV Inspection

A Novel Coarse-to-Fine Scheme for Remote Sensing Image Registration Based on SIFT and Phase Correlation

Remote Sensing Image Registration Based on Dynamic Threshold Calculation Strategy and Multiple-Feature Distance Fusion.

Method of Fusing Dual-Spectrum Low Light Level Images Based on Gray-Scale Spatial Correlation

The Processing and Fusion on Low Light Level Image and Infrared Thermal Image

Design and Implementation of A Real-Time Image Registration in an Infrared Search and Track System

Multi-source Remote Sensing Image Registration Based on Sift and Optimization of Local Self-Similarity Mutual Information

EEPNet: Efficient Edge Pixel-based Matching Network for Cross-Modal Dynamic Registration between LiDAR and Camera

Multispectral Registration Algorithm Based on Fusion of Infrared and Color Images

Multi-Band Coherent Registration Algorithm for High-Speed Target Under System Error

Real-Time Registration In Image Stitching Under The Microscope

Robust Registration of Aerial Images and LiDAR Data Using Spatial Constraints and Gabor Structural Features

Robust Registration for Ultra-Field Infrared and Visible Binocular Images

Fusion algorithm of UAV infrared image and visible image registration

Application of deep learning model based on image definition in real-time digital image fusion

A Study of Image Registration Algorithm Based on Radial-Energy Curve Method

An IR and visible image sequence automatic registration method based on optical flow