Accumulative Errors Optimization for Visual Odometry of ORB-SLAM2 Based on RGB-D Cameras

Qin, Jiangying; Li, Ming; Liao, Xiangyun; Zhong, Jiageng

doi:10.3390/ijgi8120581

Cited by 8 publications

(6 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…EPnP, UPnP (Hesch and Roumeliotis 2011;Lepetit, Moreno-Noguer, and Fua 2009) algorithms). For the mismatched points in the matching process, the Random Sampling Consensus algorithm (RANSAC (Fischler and Bolles 1981), Progressive Sampling Consensus (PROSAC) (Chum and Matas 2005)) is used to eliminate the mismatching points and accelerate the camera pose calculation (Qin et al 2019). Among them, when performing 2D-3D matching, it is necessary to search for the 3D feature points corresponding to the 2D feature points in the 3D point cloud feature library.…”

Section: Relate Workmentioning

confidence: 99%

VNLSTM-PoseNet: A novel deep ConvNet for real-time 6-DOF camera relocalization in urban streets

Qin

et al. 2021

Geo-spatial Information Science

Self Cite

View full text Add to dashboard Cite

Image-based relocalization is a renewed interest in outdoor environments, because it is an important problem with many applications. PoseNet introduces Convolutional Neural Network (CNN) for the first time to realize the real-time camera pose solution based on a single image. In order to solve the problem of precision and robustness of PoseNet and its improved algorithms in complex environment, this paper proposes and implements a new visual relocation method based on deep convolutional neural networks (VNLSTM-PoseNet). Firstly, this method directly resizes the input image without cropping to increase the receptive field of the training image. Then, the image and the corresponding pose labels are put into the improved Long Short-Term Memory based (LSTM-based) PoseNet network for training and the network is optimized by the Nadam optimizer. Finally, the trained network is used for image localization to obtain the camera pose. Experimental results on outdoor public datasets show our VNLSTM-PoseNet can lead to drastic improvements in relocalization performance compared to existing state-of-theart CNN-based methods.

show abstract

Section: Relate Workmentioning

confidence: 99%

VNLSTM-PoseNet: A novel deep ConvNet for real-time 6-DOF camera relocalization in urban streets

Qin

et al. 2021

Geo-spatial Information Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, they generally require continuous input data and occupy a large amount of the computing resources in smartphones, and they have generally been used only in a small range of VR/AR applications. In [36][37][38][39][40][41], RGB-D depth cameras were used to study high-precision real-time indoor 3D surface model reconstruction and mapping technologies. The RGB-D depth camera can provide depth information directly to the sensor while acquiring images, which improves the ability of color camera-pose estimation.…”

Section: Related Workmentioning

confidence: 99%

A Precise Indoor Visual Positioning Approach Using a Built Image Feature Database and Single User Image from Smartphone Cameras

Chen

Liao

et al. 2020

Remote Sensing

Self Cite

View full text Add to dashboard Cite

Indoor visual positioning is a key technology in a variety of indoor location services and applications. The particular spatial structures and environments of indoor spaces is a challenging scene for visual positioning. To address the existing problems of low positioning accuracy and low robustness, this paper proposes a precision single-image-based indoor visual positioning method for a smartphone. The proposed method includes three procedures: First, color sequence images of the indoor environment are collected in an experimental room, from which an indoor precise-positioning-feature database is produced, using a classic speed-up robust features (SURF) point matching strategy and the multi-image spatial forward intersection. Then, the relationships between the smartphone positioning image SURF feature points and object 3D points are obtained by an efficient similarity feature description retrieval method, in which a more reliable and correct matching point pair set is obtained, using a novel matching error elimination technology based on Hough transform voting. Finally, efficient perspective-n-point (EPnP) and bundle adjustment (BA) methods are used to calculate the intrinsic and extrinsic parameters of the positioning image, and the location of the smartphone is obtained as a result. Compared with the ground truth, results of the experiments indicate that the proposed approach can be used for indoor positioning, with an accuracy of approximately 10 cm. In addition, experiments show that the proposed method is more robust and efficient than the baseline method in a real scene. In the case where sufficient indoor textures are present, it has the potential to become a low-cost, precise, and highly available indoor positioning technology.

show abstract

“…The SLAM module estimates the current UAV pose in a map generated using the images and IMU data as illustrated in Figure 11a. ORB-SLAM2 [25,26] was used as the SLAM algorithm and was used in the RGBD mode, which utilized the left and computed depth images from the UAV to construct a map consisting of a sparse set of 3D map points while also estimating the six-degree-of-freedom (6DOF) current UAV pose in it.…”

Section: Exploration Modementioning

confidence: 99%

UAV Autonomous Localization Using Macro-Features Matching with a CAD Model

Haque

Elsaharti

Elderini

et al. 2020

Sensors

View full text Add to dashboard Cite

Research in the field of autonomous Unmanned Aerial Vehicles (UAVs) has significantly advanced in recent years, mainly due to their relevance in a large variety of commercial, industrial, and military applications. However, UAV navigation in GPS-denied environments continues to be a challenging problem that has been tackled in recent research through sensor-based approaches. This paper presents a novel offline, portable, real-time in-door UAV localization technique that relies on macro-feature detection and matching. The proposed system leverages the support of machine learning, traditional computer vision techniques, and pre-existing knowledge of the environment. The main contribution of this work is the real-time creation of a macro-feature description vector from the UAV captured images which are simultaneously matched with an offline pre-existing vector from a Computer-Aided Design (CAD) model. This results in a quick UAV localization within the CAD model. The effectiveness and accuracy of the proposed system were evaluated through simulations and experimental prototype implementation. Final results reveal the algorithm's low computational burden as well as its ease of deployment in GPS-denied environments.

show abstract

Accumulative Errors Optimization for Visual Odometry of ORB-SLAM2 Based on RGB-D Cameras

Cited by 8 publications

References 34 publications

VNLSTM-PoseNet: A novel deep ConvNet for real-time 6-DOF camera relocalization in urban streets

VNLSTM-PoseNet: A novel deep ConvNet for real-time 6-DOF camera relocalization in urban streets

A Precise Indoor Visual Positioning Approach Using a Built Image Feature Database and Single User Image from Smartphone Cameras

UAV Autonomous Localization Using Macro-Features Matching with a CAD Model

Contact Info

Product

Resources

About