Dissertation > Excellent graduate degree dissertation topics show

Research on Fiber Mask Calculation Algorithm for Cross-Sectional Image

Author: ZhangQiuHua
Tutor: WanYan
School: Donghua University
Course: Computer Software and Theory
Keywords: Edge Detection Mask edge Threshold calculating Dichotomy
CLC: TP391.41
Type: Master's thesis
Year: 2008
Downloads: 12
Quote: 0
Read: Download Dissertation

Abstract


With the continuous development of computer technology, many traditional inspection work in the textile industry gradually the computer to automatically detect replaced. The detection efficiency is improved while the stability of the detection is guaranteed. The topic for the Ministry of Education, returned overseas students funded project of the Scientific Research Foundation and the National 100 Excellent Doctoral Dissertation of special funds financed projects. Natural cellulose fibers such as cotton, linen and profiled fiber fiber computer automatic identification algorithm for automated inspection of textile fiber demand, combined with the Shanghai Entry-Exit Inspection and Quarantine Bureau. The subject was in December 2007 by the State Administration for Commodity Inspection Bureau of project identification. The edges of objects in the image is one of the most basic characteristics of the various characteristics of the fiber, the fiber cross-sectional image can be obtained through its edge information calculating. The results of edge detection of the boundary of the fiber and the structure information. Edge characterization of fiber contour information is used for feature extraction, subsequent fibers recognition algorithm in the identification judgment. Therefore, the edge information extraction correctness affect the recognition performance of the fiber. Through analysis and experimental verification, the classic image edge detection algorithm can not be well applied in fiber identification. Their output in the presence of false edges and multiple edge output. These useless ingredients affect the extraction of characteristic parameters as well as to identify the correct rate of fiber. For the characteristics of the image of the fiber cross-section, using a mask edge The method descriptions fiber outline feature; fiber mask edge calculating the adaptive threshold calculation method. Included in the calculation process of the mask edge of the two threshold values ??(referred to as the determination of the high threshold and low threshold values). Wherein the high threshold value is a global threshold, and the noise of the image, contrast, and other parameters have a direct relationship. In this article, the high threshold is used to determine the presence of fibers, and therefore the changes do not affect the the mask point size of the collection, and the presence of fibers within a certain range. The calculation results obtained by the high threshold value to determine the outline of a main portion of each of the fibers. In order to improve the fiber mask the continuity and minimize noise output, this paper is to assess the background noise of the image. Dichotomy to determine the value of the low threshold value range, so that the mask contained in the output of each of the fibers of all the edge information. Meanwhile, the output is inevitably contains a lot of noise. Presents different characteristics must be such that the edge of the noise output and fiber output low threshold determined by dichotomy, the output is a continuous fiber edge mask, while the noise Since the the Irrelevant performance isolated dot output. In this paper, this feature the fiber mask edge information extraction. Dual-threshold calculation process is the calculation of spatial connectivity, the general use of recursive program, will take up a lot of stack space. This problem can be avoided by using the connected components analysis. Experiments proved that dual-threshold algorithm obtained through the mask edge of the mask edge can be well suppressed pseudo edge and multiple edge outputs, reducing the interference component in the output of the fiber contour information, so that the final contour of the fiber obtained information with a high signal-to-noise ratio and good continuity. For severe adhesion of fibers, the edges of individual fiber obtained by fiber separation mask corresponding deformation extracted affect fiber characteristics, such as the perimeter of the fiber, an area, roundness, etc.. Hilditch thinning algorithm, which interfere with the performance of the closed curve burr. Deburred by smoothing algorithm to eliminate the the closed curve interference, mask to restore and ensure the correct operation of the subsequent algorithm of characteristic parameters such as the area and perimeter of the fiber.

Related Dissertations

  1. The Tracking Research on the Half Attribute of "Water" Culture,G122
  2. Research on Noise Reducation and Color Interpolation Algorithms Based on Bayer Pattern Images for Digital Camera,TP391.41
  3. Research on the Standard of Identify Copyright Infringement,D923.41
  4. Case Study on the Predicament of Idea’s Intellectural Property Protection,D923.4
  5. Research of Image Edge Detection Based on Wavelet Transform and EMD,TP391.41
  6. Thought and expression dichotomy theory and its practical application,D923.41
  7. Research of Line Characteristic Extraction Based on Beamlet,TP391.41
  8. An Extremal Problem on Anisotropic Corners in Finite Field,O153.4
  9. On Infringement Determination of Plagiarism of Online Game Works,D923.41
  10. Politics and administration dichotomy Thought,D035
  11. A Method for Partioning Even Clusters Based on Grids,TP311.13
  12. The Design of Micro Shape Error Measurement System Based on Image Processing,TP391.41
  13. The Almost Periodic Solutions of Differential Equations and Its Application,O175
  14. Several types of Almost Periodic Differential Equations Existence and uniqueness,O175
  15. From the traditional public administration to the new public administration: the American democratic constitutional system of remodeling,D771.2
  16. Range Image Segmentation Based on Curvature and Normal Vector,TP391.41
  17. Edge information based on infrared image segmentation method for color image,TP391.41
  18. Research on Medical Image Edge Detection Based on Mathematical Morphology,TP391.41
  19. An Investigation of Soft Morphological Edge-detection Algorithm Based on Fuzzy Theory,TP391.41
  20. Statistics on the Number of People Indoor Based on Digital Image Processing,TP391.41

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