• The Journal of Korea Robotics Society
• /
• v.17 no.3
• /
• pp.314-321
• /
• 2022

Depth information used in SLAM and visual odometry is essential in robotics. Depth information often obtained from sensors or learned by networks. While learning-based methods have gained popularity, they are mostly limited to RGB images. However, the limitation of RGB images occurs in visually derailed environments. Thermal cameras are in the spotlight as a way to solve these problems. Unlike RGB images, thermal images reliably perceive the environment regardless of the illumination variance but show lacking contrast and texture. This low contrast in the thermal image prohibits an algorithm from effectively learning the underlying scene details. To tackle these challenges, we propose multi-channel remapping for contrast. Our method allows a learning-based depth prediction model to have an accurate depth prediction even in low light conditions. We validate the feasibility and show that our multi-channel remapping method outperforms the existing methods both visually and quantitatively over our dataset.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.1
• /
• pp.302-320
• /
• 2017

In this paper, an innovative robust feature detection and matching strategy for visual odometry based on stereo image sequence is proposed. First, a sparse multiscale 2D local invariant feature detection and description algorithm AKAZE is adopted to extract the interest points. A robust feature matching strategy is introduced to match AKAZE descriptors. In order to remove the outliers which are mismatched features or on dynamic objects, an improved random sample consensus outlier rejection scheme is presented. Thus the proposed method can be applied to dynamic environment. Then, geometric constraints are incorporated into the motion estimation without time-consuming 3-dimensional scene reconstruction. Last, an iterated sigma point Kalman Filter is adopted to refine the motion results. The presented ego-motion scheme is applied to benchmark datasets and compared with state-of-the-art approaches with data captured on campus in a considerably cluttered environment, where the superiorities are proved.

• The Journal of Korea Robotics Society
• /
• v.11 no.4
• /
• pp.300-309
• /
• 2016

This paper proposes a method for motion estimation of consecutive cameras using 3-D straight lines. The motion estimation algorithm uses two non-parallel 3-D line correspondences to quickly establish an initial guess for the relative pose of adjacent frames, which requires less correspondences than that of current approaches requiring three correspondences when using 3-D points or 3-D planes. The estimated motion is further refined by a nonlinear optimization technique with inlier correspondences for higher accuracy. Since there is no dominant line representation in 3-D space, we simulate two line representations, which can be thought as mainly adopted methods in the field, and verify one as the best choice from the simulation results. We also propose a simple but effective 3-D line fitting algorithm considering the fact that the variance arises in the projective directions thus can be reduced to 2-D fitting problem. We provide experimental results of the proposed motion estimation system comparing with state-of-the-art algorithms using an open benchmark dataset.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.6
• /
• pp.99-104
• /
• 2021

The technology of estimating a movement trajectory using a monocular camera such as a smartphone and composing a surrounding 3D image is key not only in indoor positioning but also in the metaverse service. The most important thing in this technique is to estimate the coordinates of the moving camera center. In this paper, a new algorithm for geometrically estimating the moving distance is proposed. The coordinates of the 3D object point are obtained from the first and second photos, and the movement distance vector is obtained using the matching feature points of the first and third photos. Then, while moving the coordinates of the origin of the third camera, a position where the 3D object point and the feature point of the third picture coincide is obtained. Its possibility and accuracy were verified by applying it to actual continuous image data.

• The Journal of Korea Robotics Society
• /
• v.13 no.1
• /
• pp.26-30
• /
• 2018

Planetary global localization is necessary for long-range rover missions in which communication with command center operator is throttled due to the long distance. There has been number of researches that address this problem by exploiting and matching rover surroundings with global digital elevation maps (DEM). Using conventional methods for matching, however, is challenging due to artifacts in both DEM rendered images, and/or rover 2D images caused by DEM low resolution, rover image illumination variations and small terrain features. In this work, we use train CNN discriminator to match rover 2D image with DEM rendered images using conditional Generative Adversarial Network architecture (cGAN). We then use this discriminator to search an uncertainty bound given by visual odometry (VO) error bound to estimate rover optimal location and orientation. We demonstrate our network capability to learn to translate rover image into DEM simulated image and match them using Devon Island dataset. The experimental results show that our proposed approach achieves ~74% mean average precision.