Robustness enhancement of optical flow sensors accuracy to surface texture variations using point tracking algorithm

Takaloo, S ; Sharif University of Technology

327 Viewed
  1. Type of Document: Article
  2. DOI: 10.1109/ICRoM.2017.8466142
  3. Abstract:
  4. Novel architecture of high precision localization using optical flow sensor (OFS) combined with Iterative Point Tracking Algorithm (IPTA) is proposed. This work focuses on attenuation of OFS' sensitivity dependency on texture of surface over which sensor is moving. The aim for the design of experimental setup is to verify how much a robustness of OFS's sensitivity on various surfaces improves. In this regard, four different surfaces' texture including iron, paper, textile and granite stone is opted. Experimental results indicate that sensor's resolution via IPTA on surfaces of iron, paper, textile and granite stone respectively equal to 382, 460, 528 and 448 CPI. Optimal value of the algorithm parameters is calculated via Genetic Algorithm (GA). We show that IPTA is one of the effective algorithms that can enhance the robustness of OFS' resolution to surface's texture variations. © 2017 IEEE
  5. Keywords:
  6. Genetic algorithm ; High accuracy planar localization ; Optical flow sensor ; point tracking algorithm ; Surface texture ; Genetic algorithms ; Granite ; Iron ; Optical flows ; Robotics ; Textiles ; Tracking (position) ; Algorithm parameters ; Effective algorithms ; High-accuracy ; High-precision localization ; Novel architecture ; Optical flow sensors ; Point tracking ; Surface textures ; Iterative methods
  7. Source: 5th RSI International Conference on Robotics and Mechatronics, IcRoM 2017, 25 October 2017 through 27 October 2017 ; 2018 , Pages 406-410 ; 9781538657034 (ISBN)
  8. URL: https://ieeexplore.ieee.org/document/8466142