Defect inspection method and defect inspection apparatus

• US 6,556,291 B2
• Filed: 05/31/2001
• Issued: 04/29/2003
• Est. Priority Date: 05/31/2000
• Status: Expired due to Fees

• Current Assignee: Nidek Company Limited
• Original Assignee: Nidek Company Limited
• Inventors: Yonezawa, Eiji
• Primary Examiner(s): Font, Frank G.
• Assistant Examiner(s): Nguyen, Sang H.
• Application Number:
  US09/867,600
• Publication Number:
  US 20010048522A1
• Time in Patent Office: 698 Days
• Field of Search: 356/237.1, 356/237.2, 356/237.3, 356/237.4, 356/237.5, 356/600, 356/601, 356/73, 382/141, 382/144, 382/145, 382/149, 382/225, 395/11, 395/21-23
• US Class Current: 356/237.2
• CPC Class Codes: G01N 21/25 Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
• CPC Class Codes: G01N 21/88 Investigating the presence of flaws or contamination

Abstract

A defect inspection apparatus for inspecting a presence of a defect on an object includes: a first input unit which inputs wavelength characteristics of each of a plurality of samples with wavelength variation of an illumination light for inspection; a second input unit which inputs inspection conditions which an inspector sets for each of the samples as a teaching signal; a third input unit which inputs a wavelength characteristic of the object with the wavelength variation of the illumination light; a neural network which learns and stores a relationship between the inputted wavelength characteristic of each sample and the inputted inspection condition for each sample, and determines an inspection condition for the object based on the inputted wavelength characteristic of the object and the learned relationship; and a defect detector which detects a defect of the object based on the determined inspection condition of the object.

Claims

1. A defect inspection method of macroscopically inspecting a presence of a defect on an object, the defect inspection method comprising the steps of:

obtaining a plurality of images by serially illuminating a plurality of samples with a plurality of narrow-band light whose centers of wavelengths differ from each other, wherein inspection conditions for each of the plurality of samples are previously determined;

inputting wavelength characteristics of each of the plurality of samples obtained from luminance information of each image with respect to each center of wavelength to a neural network;

inputting the inspection conditions suitable for each of the samples as a teaching signal;

learning and storing the inputted wavelength characteristic of each sample and the inputted inspection condition for each sample correspondingly;

obtaining a plurality of images by serially illuminating an object, whose inspection condition is not determined, with the plurality of narrow-band light whose centers of wavelength differ from each other;

inputting a wavelength characteristic of the object obtained from luminance information of each image with respect to each center of wavelength to the neural network;

determining an inspection condition for the object based on the inputted wavelength characteristic of the object and the wavelength characteristic of each sample and the inspection condition for each sample which has been learned and stored; and

outputting the determined inspection condition for the object.

2. The defect inspection method according to claim 1 further comprising the step of:

learning and storing the inputted wavelength characteristic of the object and the determined inspection condition for the object correspondingly.

3. The defect inspection method according to claim 1, wherein in the learning and storing step, the wavelength characteristic of the following sample and the inspection condition for the following sample are correspondingly learned and stored referring to the relationship between the wavelength characteristic of the previous sample and the inspection condition for the previous sample.

4. The defect inspection method according to claim 1, further comprising the steps of:

executing differential processing between a reference image and the image of the object for obtaining a differential image;

binarizing the differential image for obtaining a binarization data; and

detecting the defect based on the binarization data.

5. The defect inspection method according to claim 4, wherein the inspection condition includes a threshold level referred in the binarizing step.

6. The defect inspection method according to claim 4, further comprising:

a step of filtering the differential image, the inspection condition includes a parameter for the filtering.

7. A defect inspection apparatus for macroscopically inspecting a presence of a defect on an object, the defect inspection apparatus comprising:

a first image obtaining unit for obtaining a plurality of images by serially illuminating a plurality of samples with a plurality of narrow-band light whose centers of wavelengths differ from each other, wherein inspection conditions for each of the plurality of samples are previously determined;

a neural network;

a first input unit which inputs wavelength characteristics of each of the plurality of samples obtained from luminance information of each image with respect to each center of wavelength to the neural network;

a second input unit which inputs the inspection conditions suitable for each of the samples as a teaching signal;

a second image obtaining unit for obtaining a plurality of images by serially illuminating an object, whose inspection condition is not determined, with the plurality of narrow-band light whose centers of wavelength differ from each other;

a third input unit which inputs a wavelength characteristic of the object obtained form luminance information of each image with respect to each center of wavelength to the neural network;

wherein the neural network correspondingly learns and stores the inputted wavelength characteristic of each sample and the inputted inspection condition for each sample, determines an inspection condition for the object based on the inputted wavelength characteristic of the object and the wavelength characteristic of each sample and the inspection condition for each sample which has been learned and stored, and outputs the determined inspection condition of the object.

8. The defect inspection apparatus according to claim 7, wherein the neural network further learns and stores the inputted wavelength characteristic of the object and the determined inspection condition for the object correspondingly.

9. The defect inspection apparatus according to claim 7, wherein the neural network learns and stores the wavelength characteristic of the following sample and the inspection condition for the following sample correspondingly referring to the relationship between the wavelength characteristic of the previous sample and the inspection condition for the previous sample.

10. The defect inspection apparatus according to claim 7, further comprising defect detecting means for executing differential processing between a reference image and the image of the object for obtaining a differential image, for binarizing the differential image for obtaining a binarization data, and for detecting the defect based on the binarization data.

11. The defect inspection apparatus according to claim 10, wherein the inspection condition includes a threshold level referred for binarizing.

12. The defect inspection apparatus according to claim 10, wherein the defect detecting means filters the differential image, the inspection condition includes a parameter for filtering.