Research On The Theory And Application Of Automated Defect Inspection System For Fabric

This paper is concerned with the study of automated defect inspection system for fabric. Computer vision data acquisition and image processing make it possible to detect and characterize defects. The gray images by means of reflectance and transmittance from woven fabric are acquired by line-scan camera. The method is based on textile surface analysis under high frequency and illuminant lighting source by image processing method.Designing the UI of internet applications with Visual Basic6.0, gathering the image data in the way of connecting SDK2000 VGA and CCD camera, With the Matlab toolbox: image processing toolbox, wavelet toolbox, find a best method which applies accurately and quickly to identify and segment the fabric detects by a mass comparisons. Interrelating three methods to design Auto-Detecting Fabric Defect System.In the process of getting image, apply the drive of SDK2000 VGA in applications to save image. In the process of processing image, apply the methods of 2-D wavelet, wavelet Denoising, threshold segmentation, math-morphologic. Making use of the latest COM builder to apply the *M files of Matlab in Visual Basic.