Browse Prior Art Database

Image Processing Method

IP.com Disclosure Number: IPCOM000118386D
Original Publication Date: 1997-Jan-01
Included in the Prior Art Database: 2005-Apr-01
Document File: 2 page(s) / 45K

Publishing Venue

IBM

Related People

Nakano, H: AUTHOR

Abstract

Disclosed is a method for detection of both macro and micro defects on the specimen's surface by using coefficients of two dimensional Spline Gaussian wavelet (SG wavelet) as the feature values.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 63% of the total text.

Image Processing Method

      Disclosed is a method for detection of both macro and micro
defects on the specimen's surface by using coefficients of two
dimensional Spline Gaussian wavelet (SG wavelet) as the feature
values.

      SG wavelet is made by a cross-product of spline wavelet and
Gaussian.  For example, the 3rd order SG wavelet means cross product
of the 3rd order spline wavelet and Gaussian.  In this invention, the
wavelet to be used for detection of defects is the complex valued
wavelet, which consists of the 3rd order SG wavelet as the real part
and the 2nd order SG wavelet as the imaginary part.  The complex
valued SG wavelet is called CSG wavelet here.

      The absolute value of coefficients of CSG wavelets for an input
image are utilized to detect defects such as periodic or streak
patterns.

      Let the 3rd order B-spline wavelet b_3(x,m).  Here, 'm' means
scale parameter of the wavelet.  In order to extend this wavelet to 2
dimensions, Gaussian g(y,a) is defined as follows:
     g(y,a)= exp((-y**2)/4a)/(4= overlay 'Y'sqrt(= overlay 'Y'Pi
      a)),                                 (1)
  where a is a scale parameter of the Gaussian, '= overlay 'Y'sqrt'
means square root, and '= overlay 'Y'Pi' means the ratio of the
circumference of a circle.
       Thus, the 3rd order SG wavelet h_3(x,y,m,a) is:
       h_3(x,y,m,a) = b_3(x,m) g(y,a).         (2)
       The coefficie...