A Versatile Method for Depth Data Error Estimation in RGB-D Sensors

Cabrera, Elizabeth V.; Ortiz, Luis E.; Silva, Bruno Marques Ferreira da; Clua, Esteban; Gonçalves, Luiz Marcos Garcia

doi:10.3390/s18093122

Cited by 23 publications

(15 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All experiments were run on an nVidia Jetson Xavier. The ArUco library [ 19 ] was employed for marker detection in the video sequences recorded with a calibrated Stereolabs ZED camera [ 70 , 71 ] mounted on top of a Pioneer 3AT robot, which is shown in Figure 13 . All sensors and devices were calibrated and shared the same coordinate system.…”

Section: Methodsmentioning

confidence: 99%

Smart Artificial Markers for Accurate Visual Mapping and Localization

Ortiz-Fernandez

Cabrera-Avila

Silva

et al. 2021

Sensors

View full text Add to dashboard Cite

Artificial marker mapping is a useful tool for fast camera localization estimation with a certain degree of accuracy in large indoor and outdoor environments. Nonetheless, the level of accuracy can still be enhanced to allow the creation of applications such as the new Visual Odometry and SLAM datasets, low-cost systems for robot detection and tracking, and pose estimation. In this work, we propose to improve the accuracy of map construction using artificial markers (mapping method) and camera localization within this map (localization method) by introducing a new type of artificial marker that we call the smart marker. A smart marker consists of a square fiducial planar marker and a pose measurement system (PMS) unit. With a set of smart markers distributed throughout the environment, the proposed mapping method estimates the markers’ poses from a set of calibrated images and orientation/distance measurements gathered from the PMS unit. After this, the proposed localization method can localize a monocular camera with the correct scale, directly benefiting from the improved accuracy of the mapping method. We conducted several experiments to evaluate the accuracy of the proposed methods. The results show that our approach decreases the Relative Positioning Error (RPE) by 85% in the mapping stage and Absolute Trajectory Error (ATE) by 50% for the camera localization stage in comparison with the state-of-the-art methods present in the literature.

show abstract

Section: Methodsmentioning

confidence: 99%

Smart Artificial Markers for Accurate Visual Mapping and Localization

Ortiz-Fernandez

Cabrera-Avila

Silva

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…Our approach is devised for situations that are similar to the works of Stateczny et al [35,36] where the key aspect of autonomous navigation is the need to avoid collisions with other objects, including shore structures. In this case, we are also developing a Vision System based on the ZED camera that works well for distances closer than 20 m [37,38], besides using a LIDAR. So the N-Boat is able to automatically detect obstacles and perform suitable maneuvers in a similar manner as pointed in the above references.…”

Section: Discussionmentioning

confidence: 99%

High-Level Path Planning for an Autonomous Sailboat Robot Using Q-Learning

Silva

Santos

Negreiros

et al. 2020

Sensors

View full text Add to dashboard Cite

Path planning for sailboat robots is a challenging task particularly due to the kinematics and dynamics modelling of such kinds of wind propelled boats. The problem is divided into two layers. The first one is global were a general trajectory composed of waypoints is planned, which can be done automatically based on some variables such as weather conditions or defined by hand using some human–robot interface (a ground-station). In the second local layer, at execution time, the global route should be followed by making the sailboat proceed between each pair of consecutive waypoints. Our proposal in this paper is an algorithm for the global, path generation layer, which has been developed for the N-Boat (The Sailboat Robot project), in order to compute feasible sailing routes between a start and a target point while avoiding dangerous situations such as obstacles and borders. A reinforcement learning approach (Q-Learning) is used based on a reward matrix and a set of actions that changes according to wind directions to account for the dead zone, which is the region against the wind where the sailboat can not gain velocity. Our algorithm generates straight and zigzag paths accounting for wind direction. The path generated also guarantees the sailboat safety and robustness, enabling it to sail for long periods of time, depending only on the start and target points defined for this global planning. The result is the development of a complete path planner algorithm that, together with the local planner solved in previous work, can be used to allow the final developments of an N-Boat making it a fully autonomous sailboat.

show abstract

“…The accuracy is thoroughly tested in 11,12 . Furthermore, in 10 , the ZED camera was found to be more accurate over 3 meters than structured light based sensors like Kinect v1 and v2.…”

Section: Methodsmentioning

confidence: 98%

“…Time-of-Flight (ToF) cameras, like the Kinect v2, were also discarded because of their poor performance in outdoor environments, as stated in 9 . Another key point for the camera choice was the work in 10 , where a detailed survey of Depth Data Error Estimation was carried out in the three most commonly used RGB-D Sensors: Zed Camera, Kinect v1 and Kinect v2. As the paper concludes (section 5.3), for distances greater than 3.5 m, the ZED camera obtains data with the lowest depth RMS error, which makes this depth sensor perfect for our dataset.…”

Section: Methodsmentioning

confidence: 99%

UASOL, a large-scale high-resolution outdoor stereo dataset

et al. 2019

View full text Add to dashboard Cite

In this paper, we propose a new dataset for outdoor depth estimation from single and stereo RGB images. The dataset was acquired from the point of view of a pedestrian. Currently, the most novel approaches take advantage of deep learning-based techniques, which have proven to outperform traditional state-of-the-art computer vision methods. Nonetheless, these methods require large amounts of reliable ground-truth data. Despite there already existing several datasets that could be used for depth estimation, almost none of them are outdoor-oriented from an egocentric point of view. Our dataset introduces a large number of high-definition pairs of color frames and corresponding depth maps from a human perspective. In addition, the proposed dataset also features human interaction and great variability of data, as shown in this work.

show abstract

A Versatile Method for Depth Data Error Estimation in RGB-D Sensors

Cited by 23 publications

References 44 publications

Smart Artificial Markers for Accurate Visual Mapping and Localization

Smart Artificial Markers for Accurate Visual Mapping and Localization

High-Level Path Planning for an Autonomous Sailboat Robot Using Q-Learning

UASOL, a large-scale high-resolution outdoor stereo dataset

Contact Info

Product

Resources

About