Dissertation > Excellent graduate degree dissertation topics show

Research and Application of Shape Based Image Detection Technology

Author: LiYiYue
Tutor: WangRenHuang
School: Guangdong University of Technology
Course: Detection Technology and Automation
Keywords: Machine vision Shape Characteristics Sequential images Symmetry Angles of rotation
CLC: TP391.41
Type: Master's thesis
Year: 2013
Downloads: 54
Quote: 0
Read: Download Dissertation

Abstract


In daily life, the shape of an object is an important parameter describing its information. In the field of machine vision and image measurement and control, shape is one of the most significant characteristics of images, and this paper is mainly about the research and discussion of the shape of images. As a large scale of meaningful visual information is included in the movement, in some case of dynamic detection, in order to detect the moving object and evaluate the moving parameters, it is necessary to combine and analyze the symmetry of sequential images. The focus of this research is to introduce how to enhance the correctness and instantaneity of the algorithms of automatic detection, through the processing and analysis of the multiple-frame sequential images.During the process of modern industrial mass-production and detection, as the standard for quality and productivity is getting higher and higher, traditional detecting methods, such as manual detecting, is not able to meet the needs of companies. Taking the detection of the badminton’s flight performance in the wind tunnel as application background, by studying some key problems in the process and analysis of the shape of sequential images, and through image detection of non-contact machine vision, this paper is to detect the on-line performance of badminton. The main research contents and innovation is as follow.1. By extracting the shape characteristics of sequential images using image processing technologies such as, shape segmentation, binaryzation, characteristic point locating, the characteristics can be extracted. And through geometrical distribution of images characteristics we calculate the shape symmetry measurement of two-dimensional projection image, as to obtain the parameter indicator that can objectively reflect the degree of the runout and tilting of the moving badminton. Finally measure the deflection condition of the moving badminton.2. Use the sequential images processing technologies to measure the revolving speed. During this process, several algorithms of evaluating the rotating angle of the related images are brought forward, compared and analyzed, including:algorithm about principal axis of inertia based on shape geometrical and physical meanings, the angle evaluation algorithm based on Zernike matrix, and the algorithm based on the matched rotating angle of related images. Considering the instantaneity and correctness of algorithms, we first utilize the algorithm of low order geometrical matrix to determine the approximate angle, and then capture samples from the shape area at equal intervals under the polar coordinates mapping, extract the characteristic curve function of the shape, and do accurate matching by using normalized product correlation functions, within the set-up angle scale, search the best partial matching optimal solutions to determine the rotating angle of the target object. The badminton’s rotate speed can be measured within certain range, the method is of reference value for some revolving speed measurement.3. The purpose is to design a system of detecting moving objects of sequential images based on parallel processing thinking, and come up with specific hardware structures and software resolving solutions based on multithreading technology, and use the advantages of computer’s multi-core processing platform to improve the detection efficiency of the system immensely. By measuring the parameter indicators of badminton’s runout and rotate speed in the wind tunnel, it offers possibility to classify the badminton.The specific algorithm is to achieve on-line detection by doing simulation experiments using Matlab tools and programming under the VC++software. The experimental results show that the series algorithms and thinking mentioned above is of feasibility and correctness, and therefore of practical values.

Related Dissertations

  1. FPGA/DSP Image Co-processor Technology and Ethernet Data Transmission,TP391.41
  2. Camera Calibration and Position and Pose Detecting on Vision Measurement System of PCB,TP391.41
  3. Research on the Theory of Visual Object Tracking Based on Particle Filter for Autonomous Robot,TP242
  4. Research on Detecting Optical Fiber Geometric Parameter Based on Machine Vision,TN253
  5. On Repetitive Control Strategies for SPWM Inverters,TM464
  6. Content-Based Thangka Image Retrieval Technology Research,TP391.41
  7. The Symmetry Identification Simulation and Statistical Analysis,O212.6
  8. Lie Symmetry, Simiarity Reduction of (2+1)-Dimensional Generalized Burgers Equation,O175.29
  9. Machine Vision’s Research and Application in the SMM of SMT,TP391.41
  10. Nonlinear Partial Differential Integral Method,O175.29
  11. Research and Implementation of Pedestrian Detection in Video Images,TP391.41
  12. Research and Implementation of Computer-Aided Color Matching System for Porcelain Teeth,R783
  13. Installation paste dimensional detection based on machine vision patch,TP391.41
  14. Research into Visual Orientation in the Weeding Robot Based on COCM,TP391.41
  15. Structure and Properties of Positive Solutions for an Integral Equation on Half Space,O175.5
  16. Research on Vision Detection Algorithm for Tracing Printing System,TP391.41
  17. Research on Image Denoising Algorithm in Machine Vision System,TP391.41
  18. Research on Machine Vision High-speed Positioning System Based on Linear Motor,TM33
  19. A Weld Quality Inspection System for Auto Parts Based on Machine Vision,TP391.41
  20. Research of Image Input and Lens Distortion Correction Technology on DM6437,TP391.41
  21. On the Construction of Two-direction Wavelets and Two-direction Multi-Wavelets,O174.2

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Image recognition device
© 2012 www.DissertationTopic.Net  Mobile