Dual-channel fabric fault detection method based on two-dimensional empirical mode decomposition
An empirical mode decomposition and detection method technology, applied in image analysis, image data processing, instruments, etc., can solve problems such as no boundary processing, lack of systematic comparison, and general comparison results
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0075] (1) For a fabric image with holes and defects (Fig. 2(a)), use geodesic dilation operator to find extreme points, mirror image continuation boundary processing, global thin plate spline interpolation after segmental interpolation and SD≤0.2 stop The optimized 2-D empirical mode decomposition of the criterion yields IMF1 (Fig. 2(b)), IMF2 (Fig. 2(c)), IMF3 (Fig. 2(d)) and residuals (Fig. 2(e));
[0076] (2) After merging IMF2 and IMF3 to obtain IMF2+3 (Figure 3(a)), use the μ±3σ threshold method to binarize it to obtain the grayscale detection channel result (Figure 3(b));
[0077](3) Measure the Laws texture of IMF1 using the L5L5 template and the W5W5 template to obtain the texture energy (Fig. The energy result is binarized using the μ±3σ threshold method to obtain two binarized results, and the two binarized results are combined to obtain the texture detection channel result (Figure 4(b));
[0078] (4) Merge the result of the gray level detection channel and the res...
Embodiment 2
[0080] (1) For a double-weft defect fabric image (Fig. 6(a)), use geodesic expansion operator to find extreme points, mirror image extension boundary processing, global thin plate spline interpolation after segmental interpolation and SD≤0.2 stop The optimized two-dimensional empirical mode decomposition of the criterion, IMF1, IMF2, IMF3 and residuals are obtained;
[0081] (2) After merging IMF2 and IMF3 to obtain IMF2+3, use the μ±3σ threshold method to binarize it to obtain the grayscale detection channel result (Figure 6(b));
[0082] (3) Use the L5L5 template and W5W5 template to measure the Laws texture of IMF1 to obtain the texture energy, and use the μ±3σ threshold method to perform binarization on the texture energy results of the L5L5 template and the texture energy results of the W5W5 template respectively, and obtain two binary values The results of binarization are combined to obtain the result of the texture detection channel (Fig. 6(c));
[0083] (4) Merge the...
Embodiment 3
[0085] (1) For a fabric image with dense road defects (Fig. 7(a)), use geodesic dilatation operator to find extreme points, image continuation boundary processing, global thin-plate spline interpolation after segmental interpolation and SD≤0.2 The optimized two-dimensional empirical mode decomposition of the stopping criterion yields IMF1, IMF2, IMF3 and residuals;
[0086] (2) After merging IMF2 and IMF3 to obtain IMF2+3, use the μ±3σ threshold method to binarize it to obtain the grayscale detection channel result (Figure 7(b));
[0087] (3) Use the L5L5 template and W5W5 template to measure the Laws texture of IMF1 to obtain the texture energy, and use the μ±3σ threshold method to perform binarization on the texture energy results of the L5L5 template and the texture energy results of the W5W5 template respectively, and obtain two binary values The results of binarization are combined to obtain the result of the texture detection channel (Fig. 7(c));
[0088] (4) Merge the ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com