Appearance inspection method and appearance inspection apparatus having high inspection processing speed

• US 20010028733A1
• Filed: 04/05/2001
• Published: 10/11/2001
• Est. Priority Date: 04/06/2000
• Status: Active Grant

• Current Assignee: Renesas Electronics Corporation
• Original Assignee: NEC Corporation
• Inventors: Sasaki, Yoshihiro, Nagao, Masahiko
• Granted Patent: US 6,741,734 B2
• Time in Patent Office: Days
• US Class Current: 382/149
• CPC Class Codes: G01N 21/956 Inspecting patterns on the surface of objects contactless testing of electronic circuits G01R31/308; testing currency G07D; manufacturing processes per se of semiconductor devices implementing a measuring step H01L22/10
• CPC Class Codes: G06T 2207/30148 Semiconductor; IC; Wafer
• CPC Class Codes: G06T 7/0004 Industrial image inspection
• CPC Class Codes: G06T 7/136 involving thresholding

Abstract

An appearance inspection method, includes (a), (b), (c), (d), (e), (f), (g), and (h). The (a) includes providing an image data in which an inspected sample is photographed. The (b) includes detecting a brightness of each of a plurality of image units included in the image data based on the image data. The (c) includes detecting the number of the image units being identical with each other in the brightness for each of the brightness. The (d) includes detecting, as a measured maximum number, the number that is maximum of the detected numbers as a result of the (c). The (e) includes computing the measured maximum number to determine a set maximum number. The (f) includes determining a threshold level of the brightness based on the set maximum number. The (g) includes converting the image data into a binary pattern based on the threshold level. The (h) includes detecting a defect of the inspected sample based on the binary pattern.

Claims

1. An appearance inspection method, comprising:

(a) providing an image data in which an inspected sample is photographed;

(b) detecting a brightness of each of a plurality of image units included in said image data based on said image data;

(c) detecting the number of said image units being identical with each other in said brightness for each of said brightness;

(d) detecting, as a measured maximum number, the number that is maximum of the detected numbers as a result of said (c);

(e) computing said measured maximum number to determine a set maximum number;

(f) determining a threshold level of said brightness based on said set maximum number;

(g) converting said image data into a binary pattern based on said threshold level; and

(h) detecting a defect of said inspected sample based on said binary pattern.

2. An appearance inspection method according to

claim 1

, wherein said set maximum number corresponds to a result when a noise component is removed from said measured maximum number.

3. An appearance inspection method according to

claim 1

, wherein said (e) is performed with first and second numbers, said first number corresponding to higher brightness with reference to said brightness corresponding to said measured maximum number, and said second number corresponding to lower brightness with reference to said brightness corresponding to said measured maximum number.

4. The appearance inspection method according to

claim 1

, wherein said (e) is performed with said brightness corresponding to the number identical with a subtracted value after subtracting a predetermined value from said measured maximum number.

5. The appearance inspection method according to

claim 1

, wherein a result of said (b) is represented by a histogram, and

wherein in said (e), a portion in which the number is larger than a subtracted value after subtracting a predetermined value from said measured maximum number of a waveform of said histogram, is approximated to a quadratic curve, and

wherein a peak value of said quadratic curve is detected as said set maximum number.

6. The appearance inspection method according to

claim 1

, wherein thr=said set maximum number×

a multiplication value+an offset value, and wherein thr is said threshold level of said (f).

7. The appearance inspection method according to

claim 6

, wherein at least one of said multiplication value and said offset value is different for each of a plurality of inspection items with regard to said inspected sample.

8. The appearance inspection method according to

claim 7

, wherein said inspected sample is an IC package, and said plurality of inspection items include a void inspection, a detection of a package crack and a package defect, a seal inspection, an extraneous substance on a lead inspection, and an extraneous substance between leads inspection.

9. The appearance inspection method according to

claim 1

, wherein when an inspecting target area to detect said defect is smaller than a preset value, said threshold level is not determined at said (d), (e) and (f), and a predetermined standard value is used as said threshold level.

10. The appearance inspection method according to

claim 1

, wherein said inspected sample is an IC package, a concave section is formed in said IC package, said concave section being provided for a pin to push said IC package out of a mold making industry when said IC package is molded, and

wherein when an inspecting target area to detect said defect is said concave section, said threshold level is not determined at said (d), (e) and (f), and two predetermined standard values are used as said threshold level, and

wherein said converting of said (g) is performed two time by using said two predetermined standard values as said threshold level, respectively to produce two said binary patterns, and

wherein said (h) includes detecting different kind of said defect with each other based on each of said two binary patterns, respectively.

11. The appearance inspection method according to

claim 1

, wherein said inspected sample is an IC package, and said IC package includes a seal, and wherein when an inspecting target area to detect said defect is said seal, said (b) includes detecting said brightness of a seal inspection portion in which said seal is expected to exist of said image data, and wherein said (c) includes detecting the number of said image units with regard to said seal inspection portion.

12. The appearance inspection method according to

claim 1

, wherein said inspected sample is an IC package, and said IC package includes a seal, and wherein a result of said (b) is represented by a histogram, and

wherein when an inspecting target area to detect said defect is said seal, said (f) includes scanning a waveform of said histogram from said set maximum number as a beginning point in direction to higher brightness, and

wherein said (f) includes detecting a peak value of a first upward convex curve next to a second upward convex curve including said set maximum number of said waveform as a result of said scanning, and

wherein thr=(said set maximum number+said peak value)/2, and wherein thr is said threshold level of said (f).

13. The appearance inspection method according to

claim 1

, wherein said inspected sample is an IC package, and said IC package includes a seal, and

wherein a result of said (b) is represented by a histogram, and

wherein when an inspecting target area to detect said defect is said seal, said (f) includes scanning a waveform of said histogram from said set maximum number as a beginning point in direction to higher brightness, and

wherein said (f) includes detecting a peak value of a first upward convex curve next to a downward concave portion of a second upward convex curve including said set maximum number of said waveform as a result of said scanning, and wherein thr=(said set maximum number+said peak value)/2, and wherein thr is said threshold level of said (f).

14. The appearance inspection method according to

claim 1

, wherein said inspected sample is an IC package, and said IC package includes a seal, and

wherein a result of said (b) is represented by a histogram, and

wherein when an inspecting target area to detect said defect is said seal, said (f) includes setting a range corresponding to a predetermined brightness of said histogram as a seal inspection portion, and

wherein a peak value in said seal inspection portion of a waveform of said histogram is detected, and

wherein thr=(said set maximum number+said peak value)/2, and wherein thr is said threshold level of said (f).

15. The appearance inspection method according to

claim 1

, wherein said inspected sample is an IC package, and said IC package includes a seal, and

wherein a result of said (b) is represented by a histogram, and

wherein when an inspecting target area to detect said defect is said seal, said (f) includes setting a range in which Es is the lowest point in said brightness of said histogram as a seal inspection portion, and

wherein said Es=said set maximum number×

a set multiplication value+a set offset value, and

wherein a peak value in said seal inspection portion of a waveform of said histogram is detected, and

wherein thr=(said set maximum number+said peak value)/2, and wherein thr is said threshold level of said (f).

16. An appearance inspection apparatus, comprising:

a camera photographing an inspected sample to produce an image data of said inspected sample;

a threshold level providing section providing a threshold level;

a binary converting section converting said image data into a binary pattern based on said threshold level; and

a judging section judging whether said inspected sample is passed or failed based on said binary pattern, and

wherein said threshold level providing section detects a brightness of each of a plurality of image units included in said image data based on said image data, and detects the number of said image units being identical with each other in said brightness for each of said brightness, and detects, as a measured maximum number, the number that is maximum of the detected numbers, and computes said measured maximum number to determine a set maximum number, and provides said threshold level based on said set maximum number.

17. An appearance inspection apparatus, comprising:

an image processing library storing a plurality of image processing items;

an inspection library storing a plurality of inspection item data, in which said plurality of image processing items are selected arbitrarily and in which said selected image processing items are performed in an arbitrary turn, and

wherein a binary converting process is included in said plurality of image processing items, and

wherein said binary converting process includes detecting a brightness of each of a plurality of image units included in an image data in which an inspected sample is photographed based on said image data, when providing a threshold level used in said binary converting process, and detecting the number of said image units being identical with each other in said brightness for each of said brightness, and detecting, as a measured maximum number, the number that is maximum of the detected numbers, and computing said measured maximum number to determine a set maximum number, and providing said threshold level based on said set maximum number.

18. The appearance inspection apparatus according to

claim 17

, further comprising:

an inspection data for each kind library storing a plurality of inspection data for each kind for an inspection target product, in which said plurality of inspection item data are selected arbitrarily in which said selected inspection item data are performed in an arbitrary turn.

19. The appearance inspection apparatus according to

claim 17

, wherein a plurality of said inspection item data includes data for a void inspection, data for a seal inspection, data for a mold loss inspection, data for an IC package direction difference detection and data for a lead curve detection.

20. The appearance inspection apparatus according to

claim 17

, wherein a plurality of said inspection item data includes a parameter used in the computation included in said binary converting process.

21. A computer readable recording medium for recording a program for a process, comprising:

(a) providing an image data in which an inspected sample is photographed;

(b) detecting a brightness of each of a plurality of image units included in said image data based on said image data;

(c) detecting the number of said image units being identical with each other in said brightness for each of said brightness;

(d) detecting, as a measured maximum number, the number that is maximum of the detected numbers as a result of said (c);

(e) computing said measured maximum number to determine a set maximum number;

(f) determining a threshold level of said brightness based on said set maximum number;

(g) converting said image data into a binary pattern based on said threshold level; and

(h) detecting a defect of said inspected sample based on said binary pattern.