FlowNorm: A Learning-based Method for Increasing Convergence Range of Direct Alignment

10/16/2019
by   Ke Wang, et al.
0

Many approaches have been proposed to estimate camera poses by directly minimizing photometric error. However, due to the non-convex property of direct alignment, proper initialization is still required for these methods. Many robust norms (e.g. Huber norm) have been proposed to deal with the outlier terms caused by incorrect initializations. These robust norms are solely defined on the magnitude of each error term. In this paper, we propose a novel robust norm, named FlowNorm, that exploits the information from both the local error term and the global image registration information. While the local information is defined on patch alignments, the global information is estimated using a learning-based network. Using both the local and global information, we achieve an unprecedented convergence range in which images can be aligned given large view angle changes or small overlaps. We further demonstrate the usability of the proposed robust norm by integrating it into the direct methods DSO and BA-Net, and generate more robust and accurate results in real-time.

READ FULL TEXT
research
05/21/2021

Direct Simultaneous Multi-Image Registration

This paper presents a novel algorithm that registers a collection of mon...
research
05/22/2019

Outlier Robust Extreme Learning Machine for Multi-Target Regression

The popularity of algorithms based on Extreme Learning Machine (ELM), wh...
research
09/25/2019

Robust Semi-Direct Monocular Visual Odometry Using Edge and Illumination-Robust Cost

In this work, we propose a monocular semi-direct visual odometry framewo...
research
10/28/2021

A recursive robust filtering approach for 3D registration

This work presents a new recursive robust filtering approach for feature...
research
05/28/2020

Local convergence of the FEM for the integral fractional Laplacian

We provide for first order discretizations of the integral fractional La...
research
09/02/2014

Effective Spectral Unmixing via Robust Representation and Learning-based Sparsity

Hyperspectral unmixing (HU) plays a fundamental role in a wide range of ...
research
12/26/2018

RegNet: Learning the Optimization of Direct Image-to-Image Pose Registration

Direct image-to-image alignment that relies on the optimization of photo...

Please sign up or login with your details

Forgot password? Click here to reset