| 摘要 |
PURPOSE:To improve the reliability of arithmetic processing for position detection even when a detection signal has low contrast by processing a corrected signal for the position detection. CONSTITUTION:An equation 1 holds, where h(x) is the intensity distribution of an image of a point pattern on a wafer which is formed through the optical pattern detection system including a reduction lens 4 and an optical enlargement system 9, f(x) is the an intensity distribution indicating the light-dark level of the pattern itself of the wafer, and g(x) is the intensity distribution of the detection signal 24. A function f'(x) shown by an equation 2 close to the f(x) is calculated by using some filter function t(x) to correct the detection signal and then a decrease in the contrast depending upon MTF characteristics of the optical detection system is recovered. For the purpose, the t(x) is calculated and stored in a memory 17 and a weighted mean shown by an equation 3 equivalent to the equation 2 is calculated for a signal g(x) stored in a memory 15 by using a product sum arithmetic part 18 and an address control part 19 and stored in a memory 21, obtaining a detection signal 25 which is made sharp.
|
