Bh. Khalaj et al., PATTERNED WAFER INSPECTION BY HIGH-RESOLUTION SPECTRAL ESTIMATION TECHNIQUES, Machine vision and applications, 7(3), 1994, pp. 178-185
A new self-reference signal processing technique is proposed for detec
ting the location of irregularities and defects in a periodic two-dime
nsional signal or image. Using high-resolution spectral estimation alg
orithms, the proposed technique first extracts the period and structur
e of repeated patterns from the image. Then a defect-free reference im
age for comparison with the actual image is produced. Since the techni
que acquires all the information needed from a single image (in contra
st to most existing methods), there is no need for a database image, a
scaling or alignment procedure or any a priori knowledge about the re
petition period of the patterns. Potential application fields for the
proposed method range from the area of wafer and mask defect inspectio
n, which includes inspection of memory chips, shift registers, switche
d capacitors, CCD arrays, and LCD displays to other areas that deal wi
th repeated structures, such as crystallography. Some results of apply
ing the proposed technique to real images from microlithography are pr
esented.