Precise Disparity Estimation for Narrow Baseline Stereo Based on Multiscale Superpixels and Phase Correlation

Abstract

<p><strong>Abstract.</strong> With the rapid development of subpixel matching algorithms, the estimation of image shifts with an accuracy of higher than 0.05 pixels is achieved, which makes the narrow baseline stereovision possible. Based on the subpixel matching algorithm using the robust phase correlation (PC), in this work, we present a novel hierarchical and adaptive disparity estimation scheme for narrow baseline stereo, which consists of three main steps: image coregistration, pixel-level disparity estimation, and subpixel refinement. The Fourier-Mellin transform with subpixel PC is used to co-register two input images. Then, the pixel-level disparities are estimated in an iterative manner, which is achieved through multiscale superpixels. The pixel-level PC is performed with the window sizes and locations adaptively determined according to superpixels, with the disparity values calcualted. Fast weighted median filtering based on edge-aware filter is adopted to refine the disparity results. At last, the accurate disparities are calculated via a robust subpixel PC method. The combination of multiscale superpixel hierarchy, adaptive determination of the window size and location of correlation, fast weighted median filtering and subpixel PC make the proposed scheme be able to overcome the issues of either low-texture areas or fattening effect. Experimental results on a pair of UAV images and the comparison with the fixed-window PC methods, the iterative scheme with fixed variation strategy, and a sophisticated implementation using global optimization demonstrate the superiority and reliability of the proposed scheme.</p>

Area-Based Dense Image Matching with Subpixel Accuracy for Remote Sensing Applications: Practical Analysis and Comparative Study

“…The patch size will not affect the presence of systematic errors [28]. Several adaptive strategies for adjusting patch size have been developed based on local structural statistics [14,76] or hierarchical variation [34,77]. The performance of various matching algorithms with varying patch sizes or adaptive strategies will be investigated in future work.…”

“…The intensity interpolation and phase-based methods are widely used in surface dynamics, monitoring for motion tracking [13,27], and local optimization-based methods are frequently used in digital image correlation for deformation measurement [4]. Therefore, most of the existing performance evaluation and error analysis of subpixel dense matching merely involve certain types of algorithms within the context of a specific application [13,21,[27][28][29][30][31][32][33][34]. In addition, there inevitably exist systematic errors induced in subpixel matching that deteriorate the application potentials [35][36][37][38].…”

Rethinking the Fourier-Mellin Transform: Multiple Depths in the Camera’s View

“…Due to FMT's robustness and high accuracy, it has been successfully applied in multiple applications, such as image registration [11,24,25], fingerprint image hashing [26], visual homing [27], point cloud registration [28], 3D modeling [29], remote sensing [12,30], and localization and mapping [31,32]. However, it requires that the capture device doesn't roll or pitch and that the environment is planar and parallel to the imaging plane.…”