Dissertation > Excellent graduate degree dissertation topics show

Visual recognition of robot soccer

Author: CaoAiZeng
Tutor: ZhengMingChun
School: Shandong Normal University
Course: Management Science and Engineering
Keywords: Robot Soccer Visual Identity Binary search algorithm BP artificial neural network Geometric transformation
CLC: TP242
Type: Master's thesis
Year: 2004
Downloads: 316
Quote: 3
Read: Download Dissertation

Abstract


Robot Soccer is a very complex high-tech combat activities, the visual system as a whole system is the only channel to obtain information, and the control system is the basis for decision-making. Based on a small group of family ROBOCUP robot soccer system, the visual system made a more in-depth research, we propose can quickly and accurately identify the ball and robot car algorithm. We present an image processing method for image segmentation and recognition in the study, while the actual system for robot soccer, robot soccer system based on the visual part of the characteristics of the technology, based on previous research, put forward the corresponding improved method herein, this thesis completed the following some work. First, a lot of reading in the foreign language references to the existing image segmentation, feature extraction, object recognition and visual robot soccer system identification method for a more comprehensive and systematic understanding. Secondly, in the soccer robot vision system, the color information is the most important basis for the identification division. Robot soccer system according to the actual situation in the visual system, analysis and comparison of the characteristics of different color space, decided to adopt closer to the human eye perception of color in the YUV color space as a work space vision system, the use of UV component YUV space and light-independent characteristics, weakening illumination changes on the visual system in the impact of the recognition process. And on this basis, the robot soccer system for real-time vision system, the characteristics of high accuracy requirements, the use of BP artificial neural networks for image color segmentation, a calculation can determine its color category, but also can improve the accuracy of classification sex. Our work is focused on research, through the existing system of robot soccer vision system identification algorithms and image processing object segmentation and recognition research, analyzes some drawbacks as well as improved methods can be proposed start of the race, based dichotomy seed point target grid search method, and in the game is running, using a fan-shaped grid based prejudge dichotomy seed point target discovery method, and the area determined in the car, filed under car color top position and color characteristics, accurate identification of a small fleet of gender and uniform number and orientation of a recognition method. The application shows that the method can accurately identify the target. Finally, the image recognition of the target position, the decision module in order to transmit the position information of the actual site, and in order to address the requirements of real-time characteristics of high, raised, not the image recognition pretreatment, and the geometric transformation Add to subsequently identify the target done to reduce the amount of calculation.

Related Dissertations

  1. Humanoid Robot Soccer System Based on Global Vision System,TP242.6
  2. Research on the Inteligent System of High Performance Concrete Mix Design in Zhujiang Triangle Area,TU528
  3. Ketones 3D-QSPR study,O641
  4. Research on Brand Image Design of Chinese Style Catering in Suzhou Area,F719.3
  5. The Application Analysis of Visual Identity System in University Anniversary,J524
  6. Research on Robocup Simulation System and Program Design Based on Reinforcement Learning,TP242
  7. Museum of Cultural Research and Application of Visual Identity Design,J525
  8. Research and Development of the Vision Sub-system of Soccer Robot,TP242
  9. Detection Method and Research for the Quality of Edible Oil Based on Near Infrared Spectroscopy,TN219
  10. Rural financial ecological environment and financial efficiency Correlation,F205;F224
  11. The AI-RCJ virtual robot soccer robot teaching in the primary school,TP242
  12. Research on Humanoid Robot Vision Systems and Motion Planning System of Shooting,TP242
  13. Research and Application of Arttificial Intelligence Technology in Game Development,TP18
  14. The soccer robotics research based on particle swarm optimization,TP242.3
  15. Research on Strategy of Robot Soccer Based on Opponent Information,TP242
  16. Study on Humanoid Soccer Robot Motion Planning,TP242.3
  17. ROBOCUP medium-sized group of soccer robots defensive strategy research,TP242
  18. Based on a global vision of the humanoid robot soccer system,TP242
  19. Path and Behavior Planning for Intelligent Robot Soccer,TP242
  20. Research and Control of Humanoid Football Robot,TP242
  21. Intelligent mobile robot and robot soccer Strategy,TP242

CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Robotics > Robot
© 2012 www.DissertationTopic.Net  Mobile