ABSTRACT
This paper proposes a camera pose estimation algorithm based on deep neural network, which returns the translation and rotation of the camera based on supervised deep learning. This paper uses the ORB algorithm to extract the feature points of the image, and the feature points are labeled on the color image for training and testing. The deep neural network based on the structure of recurrent convolution neural network (RCNN). Firstly, some features extracted by the convolution neural network. Then,it builds the order model base on the RCNN network. The RMS error is used as the loss function to train the network, in which the rotation is expressed by Euler angle. Finally,experiments on KITTI VO dataset show that the proposed method is effective.
- Nister D, Naroditsky O, Bergen J. Visual odometry {C}Computer Vision and Pattern Recognition. 2004:I--1Google Scholar
- Bailey T, Durrant-whyte H. Simultaneous localization and mapping (slam) {J}. Robotics & Automation Mag- zine,2006,13(3):108--117.Google Scholar
- Lowe D G. Distinctive image features from scale-invariant keypoints{J}.International Journal of Computer Vision,2004,60(2):91--110. Google ScholarDigital Library
- Bay H, Tuytelaars T, van Gool L. SURF: Speeded up robust features{C}//9th European Conference on Computer Vision. Berlin, Germany: Springer, 2006:404--417. Google ScholarDigital Library
- Rublee E, Rabaud V, Konolige K, et.al. ORB:An efficient alternative to SIFT or SURF{C} //IEEE, 2011:2564--2571. Google ScholarDigital Library
- Klein G. Parallel tracking and mapping for small AR workspaces{C}// Proc of the 6th IEEE and ACM International Symposium on Mixed and Augmented Reality (ISMAR 2007). New York, NY:ACM Press, 2007:225--234. Google ScholarDigital Library
- Mur-artal R, Montiel J M M, Tardos J D. ORB-SLAM: a versatile and accurate monocular SLAM system{J}. IEEE Transactions on Robotics, 2015, 31(5):1147--1163.Google ScholarDigital Library
- Engel J, Schps T, Cremers D. LSD-SLAM: Large-scale direct monocular SLAM{C}//European Conference on Computer Vision, Cham;Springer Press,2014:834--849.Google Scholar
- Li Ruihao, Wang Sen, Long Zhiqiang, et al. UnDeepVO: Monocular Visual Odometry through Unsupervised Deep Learning{C}// arXiv.arXiv preprint:1709.06841, 2017.Google Scholar
- Kendall A, Grimes M, Cipolla R. PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization{C}// IEEE International Conference on Computer Vision. Piscataway, NJ: IEEE Press,2015:2938--2946. Google ScholarDigital Library
- Costante G, Mancini M, Valigi P, et al. Exploring Representation Learning With CNNs for Frame-to-Frame Ego-Motion Estimation{J}. IEEE Robotics & Automation Letters, 2015, 1(1):18--25.Google ScholarCross Ref
- Wang S, Clark R, Wen H, et al. DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent Convolutional Neural Networks{C}// 2017 IEEE International Conference on Robotics and Automation (ICRA). Piscataway, NJ: IEEE Press, 2017.Google Scholar
Index Terms
- Camera Pose Estimation Method Based on Deep Neural Network
Recommendations
A hybrid deep learning model by combining convolutional neural network and recurrent neural network to detect forest fire
AbstractForest fire poses a serious threat to wildlife, environment, and all mankind. This threat has prompted the development of various intelligent and computer vision based systems to detect forest fire. This article proposes a novel hybrid deep ...
Scene text recognition using residual convolutional recurrent neural network
Text is a significant tool for human communication, and text recognition in scene images becomes more and more important. In this paper, we propose a residual convolutional recurrent neural network for solving the task of scene text recognition. The ...
Sentiment Analysis of Network Comments Based on GCNN
CSAI '18: Proceedings of the 2018 2nd International Conference on Computer Science and Artificial IntelligenceRecently, satisfactory performance has been achieved on sentiment analysis tasks by using recurrent neural network (RNN). As a derived model of RNN, gated recurrent units (GRU) model has a great advantage in dealing with long text sequence problems, ...
Comments