Feature detection and matching in images with radial distortion

Abstract

Image keypoints are broadly used in robotics for different purposes, ranging from recognition to 3D reconstruction, passing by SLAM and visual servoing. Robust keypoint matching across different views is problematic because of the relative motion between camera and scene that causes significant changes in feature appearance. The problem can be partially overcome by using state-of-the-art methods for keypoint detection and matching, that are resilient to common affine transformations such as changes in scale and rotation. Unfortunately, these approaches are not invariant to the radial distortion present in images acquired by cameras with wide field-of-view. This article proposes modifications to the Scale Invariant Feature Transform (SIFT), that improve the repeatability of detection and effectiveness of matching in the presence of distortion, while preserving the characteristics of invariance to scale and rotation. These modifications require an approximate modeling of the image distortion, and consist in using adaptative gaussian filtering for detection and implicit gradient correction for description. Extensive experiments, with both synthetic and real images, show that our method outperforms explicit distortion correction using image rectification.

A Spherical Model Based Keypoint Descriptor and Matching Algorithm for Omnidirectional Images

“…In addition, the pSIFT and sSIFT methods may produce false keypoints because of the interpolation artifacts. To avoid interpolation, Lourenço et al proposed the SIFT with radial distortion (sRD-SIFT) algorithm [8,9] based on the adaptive Gaussian filter and the compensatory gradient operator, by referring to the distortion model of Fitzgibbon [10] and the SIFT descriptor. Although the sRD-SIFT algorithm improves the matching effect of images with small distortion, it is worse 2 Advances in Mechanical Engineering than SIFT in the cases of large distortion.…”

Scale space for central catadioptric systems: Towards a generic camera feature extractor

“…Then, the approximate spherical diffusion is defined as the convolution of the stereographic image with the stereographic version of the Gaussian kernel. More recently [14] propose an improvement to the SIFT detector by introducing radial distortion into the scale-space computation. In [5] a framework to perform scale space analysis for omnidirectional images using partial differential equations is proposed.…”

Matching cylindrical panorama sequences using planar reprojections

“…It was also studied in the case of images with radial distortion [6], wide angle images [4] and spherical panoramas [15]. The strategy used in [3,9,15] is to work with reprojections of the images to be matched.…”

“…The strategy used in [3,9,15] is to work with reprojections of the images to be matched. As for [4,6], they rather modify the convolution step in the definition of the SIFT features so that it applies on wide angles images [4] or on images with radial distortion [6].…”

“…This is also the case with planar images and this is one of the reasons why ASIFT [10] was introduced by Morel and Yu. The methods presented in [3,4,6,15] do not take this fact into account. Mauthner et al [9] uses local affine frame to compensate for perspective changes.…”