Dissertation > Excellent graduate degree dissertation topics show

Fabric Detect Detection Method Based on Texture Gradient

Author: ShenJing
Tutor: YangXueZhi
School: Hefei University of Technology
Course: Signal and Information Processing
Keywords: fabric defect inspection texture gradient watershed segmentation texture enhancemence Markov random field model
CLC: TP274.4
Type: Master's thesis
Year: 2009
Downloads: 162
Quote: 2
Read: Download Dissertation

Abstract


In the textile industry, quality control is of vital importance for fabric products. Currently, the inspection task is primarily performed by human inspectors and hence heavily relies on their experience, judgement and attention. Automated inspection of the fabric defects is becoming an attractive alternative to the human visual inspection in modern textile industry. The main goal of fabric automated inspection is able to precisely locate the defect regions in the fabric. In recent years, texture gradientis a popular tool for image segmentation that considerally improves contour detection performance, and it is further used in the inspection of fabric defects. The main work is as follows:1. To develop a texture image segmentation method based on texture gradient. Based on texture gradient and marker-based watershed, simple texture watershed is designed for texture image segmentation. In order to reduce noise (which in a general sense includes fine texture and non-stationary behaviors), nonlocal means (NL-means) filter is used for enhancing the differences between the defect texture and background texture. Combined with the texture watershed transform, it is called texture watershed based on texture enhancemence. The defect regions in fabric images can be defected accurately in this method.2. To develop a method for texture defect detection based on texture gradient and MRF. By analyzing the model of MRF, the automated texture defect detection based texture gradient and MRF is proposed to overcome the drawback of simple texture watershed. In this method, Markov random field model is used in the final segmentation, followed by the texture enhancemence and texture gradient.3. To develop a method for fabric defect detection based on texture gradient. The three methods based on texture gradient are used in the inspection of fabric defects. Experiment results have demonstrated that simple texture watershed can get fast and accurate result of fabric image containing obviously different texture. The texture watershed based on texture enhancement can get accurate result of more difficultly defected fabric image, but it costs more time. The detection based on texture gradient and MRF model can get accurate result automatically.

Related Dissertations

  1. Research on Medical Image Segmentation Method Based on Markov Random Field Model,TP391.41
  2. Active Contour Models Based on Texture Subspace Components for Image Segmentation,TP391.41
  3. Research and Realization of Boundary Detecting Methods in Natural Images,TP391.41
  4. The Research of Image Line Drawings Based on Conditional Random Fields,TP391.41
  5. Depth Perception in Cue Conflict and the Mechanism of Formation,TP391.41
  6. Bark Texture Synthesis by Patch-Based Sampling,TP391.41
  7. 2D Shape Blending Based on Visual Feature Decomposition and Its Applications,TP391.4
  8. The Study of the Tracking Algorithm of the Neural Stem Cells,R318
  9. Segmentation of SAR Image Based on Mixture Multiscale ARMA Model,TP391.41
  10. Background clutter suppression and dim target detection technology,TP274.4
  11. Researches on Technologies of Content-Based Multi-Hierarchy Semantic Video Object Description Extraction,TP391.41
  12. Study on Video Moving Objects Detection and Real-Time Processing System,TN919.81
  13. Image Foreground Extraction Based on Graph Cut,TP391.41
  14. Rock Fragmentation Analysis Based on Digital Image Processing Technology,TP391.41
  15. Research and Implementation of K- means clustering of remote sensing images and watershed segmentation algorithm,TP751
  16. Breast Mass Segmentation in Digitized Mammograms Based on Watershed and Level Set,TP391.41
  17. Research on Individual Tree Identification and Crown Segmentation Algorithm in High Spatial Resolution Remote Sensing Imagery,TP391.41
  18. Research on Segmentation of Cigarettes Image Based on Mathematical Morphology,TP391.41
  19. CT image sequences based on three-dimensional reconstruction of vascular structures Method,TP391.41
  20. The Research and Improvement of Two Image Segmentation Methods Based on CT Images,TP391.41
  21. Adhesion particle image analysis technology based on watershed segmentation,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Automation systems > Data processing, data processing system > Centralized testing and roving detection system
© 2012 www.DissertationTopic.Net  Mobile