Pattern shape evaluation method

1. A pattern shape evaluation method, comprising steps of:

generating a tolerance area of a normal pattern of an electronic device according to a plurality of lines constituting the normal pattern for each of a plurality of directions, in accordance with: pattern data based on design data of the electronic device, pattern data based on a photographed image or pattern data based on a process simulation;

extracting contour data of a pattern of the electronic device from an image obtained by photographing a circuit pattern of the electronic device; and

detecting a portion of the contour data as representing a defect of the pattern of the electronic device, when the portion of the contour data corresponding to the tolerance area does not satisfy a predetermined direction condition relative to the tolerance area of the normal pattern.

2. The pattern shape evaluation method according to claim 1 , wherein

when the contour data of the pattern is extracted from the image of the circuit pattern, contour pixels of the pattern are detected from the image of the circuit pattern in a pixel unit or a subpixel unit, and

the contour data include new contour pixels obtained through an approximation process for adjacent contour pixels.

3. The pattern shape evaluation method according to claim 1 , wherein when the defect of the pattern is to be detected, a portion of pattern direction data of the contour data outside a predetermined range of angle is judged as a defect of the pattern.

4. The pattern shape evaluation method according to claim 3 , wherein

pattern direction data related to the normal pattern include pattern direction range data, and

when the defect of the pattern is to be detected and when the pattern direction data of the contour data, representing an area of the pattern, is outside a pattern direction range defined by the pattern direction range data, the area of the pattern is judged as a defect area.

5. The pattern shape evaluation method according to claim 4 , further comprising a step of judging a defect type of the defect area from a state of a particular pattern of one of the design data of the electronic device, the image obtained by photographing the circuit pattern of the electronic device, and pattern shape data obtained through a process simulation, the particular pattern being located at a position of the defect area.

6. The pattern shape evaluation method according to claim 1 , wherein

when the defect of the pattern is to be detected, a pattern position represented by the contour data is compared with pattern shape range data generated from the design data of the electronic device, and

when the pattern position represented by the contour data is outside a pattern shape range defined by the pattern shape range data, an area represented by the contour data, that is outside the pattern shape range is judged as a defect of the pattern.

7. The pattern shape evaluation method according to claim 1 , wherein a defect having a continuous contour is detected from a pattern defect, and an area containing the defect having the continuous contour is detected as a defect area.

8. The pattern shape evaluation method according to claim 1 , further comprising steps of:

detecting a contour having an opposing relation from a pattern defect, and

detecting an area including the contour having the opposing relation as a defect area.

9. A pattern shape evaluation device, comprising:

means for generating a tolerance area of a normal pattern of an electronic device according to a plurality of lines constituting the normal pattern for each of a plurality of directions, in accordance with: pattern data based on design data of the electronic device, pattern data based on a photographed image or pattern data based on a process simulation;

means for extracting contour data of a pattern of the electronic device from an image obtained by photographing a circuit pattern of the electronic device; and

means for detecting a portion of the contour data as representing a defect of the pattern of the electronic device, when the portion of the contour data corresponding to the tolerance area does not satisfy a predetermined direction condition relative to the tolerance area of the normal pattern.

10. A semiconductor inspection system, comprising:

the pattern shape evaluation device according to claim 9 ; and

a computer equipped with a scanning electron microscope and control means for the scanning electron microscope.

11. The pattern shape evaluation device according to claim 9 , wherein the means for detecting a defect of the pattern includes means for detecting a discontinuous area represented by the contour data as a defect area of the pattern, through comparison of pattern shape range data generated from the design data of the electronic device, pattern direction data related to the normal pattern and the contour data.

12. The pattern shape evaluation device according to claim 11 , wherein the means for detecting a defect of the pattern includes means for measuring a length of the discontinuous area, and judges a further defect of the pattern through comparison of the measured length with a threshold value of a defect size designated by a user.

13. The pattern shape evaluation device according to claim 9 , further comprising means for generating measuring sequence data for measuring a pattern area corresponding to the detected defect, in accordance with defect data representing the detected defect.

14. The pattern shape evaluation device according to claim 13 , wherein the measuring sequence data generating means determines a scan direction of a particular image to be used for measurement, in accordance with:

coordinate information of the defect data, or

shape information on pattern shape data extracted from the design data and representing an area adjacent to the detected defect or extracted from an image obtained by photographing the circuit pattern of the electronic device, or

shape information on pattern shape data obtained through a process simulation.

15. The pattern shape evaluation device according to claim 13 , wherein the measuring sequence data generating means determines two measuring points including coordinate data of the defect data in accordance with:

coordinate information of the defect data, or

shape information on pattern shape data extracted from the design data and representing an area adjacent to the detected defect or extracted from an image obtained by photographing the circuit pattern of the electronic device, or

shape information on pattern shape data obtained through a process simulation.

16. The pattern shape evaluation device according to claim 13 , wherein the measuring sequence data generating means determines two measuring points including coordinate data of the defect data and determines a rectangle area, a rectangle size, and a line integration direction for generating a line profile at each of the two measurement points through line integration, in accordance with:

coordinate information of the defect data, or

shape information on pattern shape data extracted from the design data and representing an area adjacent to the detected defect or extracted from an image obtained by photographing the circuit pattern of the electronic device, or

shape information on pattern shape data obtained through process simulation.

17. The pattern shape evaluation device according to claim 16 , wherein the rectangle size is determined from a pattern shape of the defect data.

18. The pattern shape evaluation device according to claim 13 , wherein in accordance with the measuring sequence data, the measuring sequence data generating means photographs an image of a defect area, measures the defect area, and measures a distance between the defect area and a closest pattern represented by pattern shape data extracted from the design data or an image obtained by photographing the circuit pattern of the electronic device or pattern shape data obtained through a process simulation.

19. The pattern shape evaluation device according to claim 13 , wherein the measuring sequence data generating means determines two measuring points on the contour data including coordinate data of the defect data and measures a distance between the two measuring points, in accordance with:

coordinate information of the defect data, or

shape information on pattern shape data extracted from the design data and representing an area adjacent to the detected defect data or extracted from an image obtained by photographing the circuit pattern of the electronic device or

shape information on pattern shape data obtained through a process simulation.