Optical flow-based branch segmentation for complex orchard environments

被引:5
|
作者
You, Alexander [1 ]
Grimm, Cindy [1 ]
Davidson, Joseph R. [1 ]
机构
[1] Oregon State Univ, Collaborat Robot & Intelligent Syst CoRIS Inst, Corvallis, OR 97331 USA
关键词
D O I
10.1109/IROS47612.2022.9982017
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Machine vision is a critical subsystem for enabling robots to be able to perform a variety of tasks in orchard environments. However, orchards are highly visually complex environments, and computer vision algorithms operating in them must be able to contend with variable lighting conditions and background noise. Past work on enabling deep learning algorithms to operate in these environments has typically required large amounts of hand-labeled data to train a deep neural network or physically controlling the conditions under which the environment is perceived. In this paper, we train a neural network system in simulation only using simulated RGB data and optical flow. This resulting neural network is able to perform foreground segmentation of branches in a busy orchard environment without additional real-world training or using any special setup or equipment beyond a standard camera. Our results show that our system is highly accurate and, when compared to a network using manually labeled RGBD data, achieves significantly more consistent and robust performance across environments that differ from the training set.
引用
收藏
页码:9180 / 9186
页数:7
相关论文
共 50 条
  • [1] An overview of optical flow-based approaches for motion segmentation
    Anthwal, Shivangi
    Ganotra, Dinesh
    IMAGING SCIENCE JOURNAL, 2019, 67 (05): : 284 - 294
  • [2] A novel optical flow-based representation for temporal video segmentation
    Akpinar, Samet
    Alpaslan, Ferdanur
    TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES, 2017, 25 (05) : 3983 - 3993
  • [3] Optical flow-based segmentation of containers for automatic code recognition
    Atienza, V
    Rodas, A
    Andreu, G
    Pérez, A
    PATTERN RECOGNITION AND DATA MINING, PT 1, PROCEEDINGS, 2005, 3686 : 636 - 645
  • [4] Moving Obstacle Segmentation with an Optical Flow-based DNN: an Implementation Case Study
    Karoly, Artur, I
    Elek, Renata Nagyne
    Haidegger, Tamas
    Galambos, Peter
    INES 2021: 2021 IEEE 25TH INTERNATIONAL CONFERENCE ON INTELLIGENT ENGINEERING SYSTEMS, 2021,
  • [5] A multiresolution flow-based multiphase image segmentation
    Barcelos, C. A. Z.
    Barcelos, E. Z.
    Cuminato, J. A.
    2009 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN AND CYBERNETICS (SMC 2009), VOLS 1-9, 2009, : 3002 - +
  • [6] Optical flow-based probabilistic tracking
    Lucena, M
    Fuertes, JM
    Gomez, JI
    de la Blanca, NP
    Garrido, A
    SEVENTH INTERNATIONAL SYMPOSIUM ON SIGNAL PROCESSING AND ITS APPLICATIONS, VOL 2, PROCEEDINGS, 2003, : 219 - 222
  • [7] Potential benefit of flow-based routing in multihomed environments
    Manilici, Vlad
    Wundsam, Andreas
    Feldmann, Anja
    Vidales, Pablo
    EUROPEAN TRANSACTIONS ON TELECOMMUNICATIONS, 2009, 20 (07): : 650 - 659
  • [8] Evolutionary Optimization Applied for Fine-Tuning Parameter Estimation in Optical Flow-based Environments
    Pereira, Danillo Roberto
    Delpiano, Jose
    Papa, Joao Paulo
    2014 27TH SIBGRAPI CONFERENCE ON GRAPHICS, PATTERNS AND IMAGES (SIBGRAPI), 2014, : 125 - 132
  • [9] Miniaturized optical chemosensor for flow-based assays
    Pokrzywnicka, Marta
    Cocovi-Solberg, David J.
    Miro, Manuel
    Cerda, Victor
    Koncki, Robert
    Tymecki, Lukasz
    ANALYTICAL AND BIOANALYTICAL CHEMISTRY, 2011, 399 (03) : 1381 - 1387
  • [10] Optical Flow-based Face Tracking in The Mummy
    Andrus, Curtis
    Balint, Endre
    Deng, Chong
    Coupe, Simon
    ACM SIGGRAPH 2017 TALKS, 2017,