JPH07229842A - Method and equipment for inspecting dust particle in ic - Google Patents

Abstract

PURPOSE:To detect a dust particle having small difference in gray level from a lead or a dust particle present between the leads by binarizing the images at the shoulder part, inclining face part, and flat part of a lead to be inspected at binary levels inherent thereto. CONSTITUTION:A camera 1 picks up the image at the part of a lead 21 in an IC 2 to be inspected and outputs an image data (a). A binarizing means 6 converts the gray image data into a binary image data (e) where the inclining face part 24 of a lead 21 has '0' level and the flat part 23 has '1' level. An X-ray projection means 7 measures the number of pixels of '1' level in the data (e) for the pixel rows in the longitudinal and vertical directions of the lead 21. A lead area division means 9 delivers a signal (h) for dividing the lead 21 into three regions of shoulder part 25, inclining face part 24 and flat part 23. A detecting means 10 divides the image data for one side based on the division signal (h). Binarization is performed with inherent binarization level for each area and a pixel of '1' is recognized as the lead 21 thus detecting 10 a dust particle.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an IC foreign substance inspection apparatus and method for detecting foreign substances such as resin scraps, fiber scraps, and solder scraps attached to IC leads.

[0002]

2. Description of the Related Art As a conventional inspection method for detecting foreign matter attached to an IC, there is an "IC lead detection method" disclosed in Japanese Patent Laid-Open No. 61-66908.

FIG. 10 is a block diagram showing this conventional inspection method. This conventional inspection method uses I
The image pickup device 31 for inputting the image of C, the frame memory 32 for storing the image input from the image pickup device 31, and the image data from the frame memory 32 are subjected to arithmetic processing and IC
When it is detected that foreign matter is attached to the lead of I
It is composed of an arithmetic processing unit 33 that outputs a C reject signal.

The detection of foreign matter attached to the IC lead by this conventional inspection method is as follows. The image data output from the frame memory 32 is input to the arithmetic processing unit 33,
The lead part is "1" and the background is "1" depending on the appropriate binarization level.
After the binarization processing is performed so that the dots become 0 ", the binarized data is" 1 "in the longitudinal direction of the lead in the area on the lead.
It is inspected whether there is a portion that changes from "0" to "1", and if there is, it is determined that foreign matter is attached to the IC lead, and a defect signal is output.

[0005]

In the above-mentioned conventional method for inspecting foreign matter adhering to an IC lead, an image of the entire lead is binarized at the same binarization level to set the lead portion to "1". Therefore, the binarization level must be set low so that the slope of the lead having a small amount of reflected light becomes "1". Therefore, when foreign matter with high transmittance such as fiber scraps or resin scraps with good reflectance adheres to the leads, the difference in gray value between the lead part and the foreign matter part of the captured image is small, so it is a binary value. It is difficult to set only the foreign matter portion to "0" in the lead portion of the digitized image, and there is a drawback that the foreign matter cannot be detected.

Further, the above-mentioned conventional inspection method has a defect that foreign matter adhered between the leads cannot be detected.

[0007]

In a foreign matter inspection apparatus for an IC according to the present invention, a lead to be inspected comprises a shoulder on a flat portion on the mold side, a flat portion on the tip side, and an inclined surface portion located in the middle thereof. A camera for picking up an image of the IC, an AD conversion unit for inputting image data from the camera, performing AD conversion, and outputting gray-scale image data, and the I-image from the gray-scale image data.
Only the inspection area cutout means for cutting out the inspection area grayscale image data in the range including all the leads for one side of C, and the flat portion and the shoulder portion of the lead where the camera receives the inspection area grayscale image data and has a large amount of reflected light. First binarizing means for binarizing so as to be "1" and binarized image data binarized by the first binarizing means are inputted and are perpendicular to the longitudinal direction of the lead. Projecting means for outputting X projection data obtained by measuring the number of pixels of "1" in each image row in the X direction, and a portion of the lead of the inspection area grayscale image data where the X projection data is smaller than a predetermined value. A portion corresponding to the central portion is a slope image, and a portion corresponding to the mold side portion of the IC larger than the predetermined value is a shoulder image, and a portion corresponding to the tip side portion of the lead larger than the predetermined value is a shoulder image. The flat part image and The lead area dividing means, the shoulder image, the slope image, and the flat image are binarized with a unique binarization level, and a pixel of "1" is identified as the lead portion. And a foreign matter detection unit for each divided area for detecting foreign matter attached to the lead.

In the IC foreign matter inspection method of the present invention, the IC to be inspected is imaged by a camera, in which the lead to be inspected is composed of a shoulder portion of a flat portion on the mold side, a flat portion on the tip side, and a slope portion located in the middle thereof. Inspection area grayscale image data in a range including all leads for one side of the IC is cut out from the AD-converted grayscale image data, and the amount of reflected light received by the camera for the inspection area grayscale image data is large and a flat portion of the lead and The X projection data obtained by measuring the number of "1" pixels in each pixel row in the X direction perpendicular to the longitudinal direction of the lead of the binarized data binarized so that only the shoulder portion becomes "1" is determined. A portion of the slope image of the portion corresponding to the center of the lead in a portion where the X projection data obtained by dividing the inspection area grayscale image data by dividing the inspection area grayscale image data is smaller than the predetermined value, and the slope image of the IC is larger than the predetermined value. The lead image is obtained by binarizing the shoulder image of the portion corresponding to the field side portion and the flat portion image of the portion corresponding to the tip end side portion of the lead that is larger than the predetermined value with a unique binarization level for each. Is detected, and the foreign matter attached to the lead is detected.

[0009]

Embodiments of the present invention will now be described with reference to the drawings.

FIG. 1 is a block diagram showing an IC foreign matter inspection apparatus according to an embodiment of the present invention.

In the figure, a camera 1 is an IC to be inspected.
The portion of the lead 21 of 2 is imaged and the analog image data a is output. The camera 1 is placed on the upper surface side of the IC 2 to be inspected, and illuminates the IC 2 with illumination (not shown) on a ring having the camera 1 as a center. The AD conversion means 3 inputs the analog image data a, performs AD conversion, and outputs grayscale image data b. The inspection area cutout unit 4 cuts out the grayscale image data of a portion including all the leads 21 for one side of the IC 2 from the grayscale image data b in a preset range,
The inspection area grayscale image data c is output. Image storage means 5
Stores the inspection area grayscale image data c and outputs this as storage data. The first binarizing means 6 sets the slope portion 24 of the lead 21 in which the grayscale image data for one side based on the storage data d is preset to “0”, and the flat portion 2 of the lead 21.
3 is converted into binary image data e by the first binarization level that makes "1" and output. The X projection means 7 measures the number of pixels of "1" of the binarized image data e for each image element row in the direction perpendicular to the longitudinal direction of the lead (hereinafter referred to as the X direction), and outputs the measurement data f.

The second binarizing means 8 inputs the measurement data f, binarizes the measurement data f at a preset second binarization level, and outputs projection binarization data g. The lead area dividing means 9 uses "0" and "1" of the projection binarized data g.
Using the continuous region of the lead 21, the lead 21 is moved from the mold 22 toward the tip of the lead 21, the shoulder portion 25, the slope portion 24, and the flat portion 2.
A lead area division signal h for dividing into three areas of 3 is output. The divided area foreign matter detection unit 10 divides the grayscale image data for one side of the storage data d by the read area division signal h, and outputs a binarization level, a lead width determination value, and an inter-lead determination value suitable for each area. Foreign matter determination is performed using this. The foreign matter detection method is binarized into binarized data for each area and "1"
The number of consecutive pixels of is compared with the upper limit value and the lower limit value of the read width judgment value, and the number of consecutive pixels of "0" is compared with the upper limit value and the lower limit value of the inter-read judgment value, and the number of consecutive pixels outside the judgment value range If there is, it is determined that foreign matter is attached.

Next, the principle of the IC foreign matter inspection method of this embodiment will be described with reference to FIGS. 2 to 9 are diagrams for explaining the principle of lead area division in the IC foreign matter inspection method of this embodiment.

FIG. 2 is a pattern diagram of the storage data d in which the grayscale image data c of the lead 21 for one side of the IC 2 is stored. FIG. 3 is a pattern diagram of the binarized image signal e, and the hatched portion is the region of “1” after binarization. Since the shoulder portion 25 of the lead 21 and the flat portion 23 have high reflectance, the first binarizing means 6 forms an area of "1". FIG. 4 is a graph of the measurement data f output from the X projection means 7, and the vertical axis shows the position in the longitudinal direction of the lead 21 in positional correspondence with FIG. The horizontal axis represents the measurement frequency, and the regions 23 'and 25' corresponding to the flat portion 23 and the shoulder portion 25 of the lead 21 have a larger value than the region 24 'to be attached to the slope portion 24. There is. The value of the measurement data f corresponding to the slope having a small reflectance is "0" or a value close to "0". FIG. 5 shows a graph of the projection binarized data g, and the regions H1 and H in which "1" of the projection binarized data are continuous.
3 corresponds to the flat portion 23 and the shoulder portion 25, respectively, and the area H
A continuous region H2 of "0" between 1 and H3 has a slope 24.
Corresponding to. Therefore, in the lead area dividing means 9, the area H1 in which "1" and "0" of the projection binarized data g are continuous,
A lead area division signal h for dividing the longitudinal direction of the lead in accordance with H2 and H3 is output. FIG. 6 is an image obtained by dividing the storage data d input to the divided area foreign matter detection means 10 by the read area division signal h. The flat portion image 26, the slope portion image 27, and the shoulder portion image 28 shown in FIG. 6 respectively correspond to the regions H1, H2, and H3 shown in FIG.

FIG. 7 is a pattern diagram in which the divided image is binarized at a binarization level preset for each area so that the lead portion of each area becomes "1". For the flat part image 26 and the shoulder part image 28 corresponding to the flat part 23 and the shoulder part 25 having a high reflectance, a binary value larger than the binarization level of the slope part image 27 corresponding to the slope part 24 having a low reflectivity. The binarization level is set so that the lead part of the binarized image becomes "1" not only in the flat part image 26 and the shoulder part image 28 but also in the slope part image 27. In addition, in the flat portion image 26 and the shoulder portion image 28, if a small binarization level is set in accordance with the slope portion image 27, the foreign matter attached to the shoulder portion 25 and the flat portion 23 is reflected by the small binarization level. Only the foreign matter having a large difference between the lead and the reflectance can be detected. However, the flat portion image 2
6. By setting the binarization level of the shoulder image 28 higher than that of the slope image 27 and setting the binarization level of the flat image 26 and the shoulder image 28 to a value as large as possible, the flat portion 23 Also, the shoulder 25 can detect foreign matter having a small difference in reflectance from the lead.

FIG. 8 is a graph of measurement data obtained by projecting the number of pixels of each divided binarized image "1" shown in FIG. 7 in the direction parallel to the longitudinal direction of the lead. 8 (a)-
(C) is a graph for the binarized image of each of the flat portion image 26, the slope portion image 27, and the shoulder portion image 28. The horizontal axis of FIG. 8 positionally corresponds to FIG. 7, and shows the position in the direction perpendicular to the longitudinal direction of the lead. The vertical axis indicates the measurement frequency, and the portion corresponding to the lead has a value larger than that between the leads. The value of the measurement data corresponding to the lead is “0” or a value close to “0”. FIG. 9 (a)
8A to 8C are graphs in which the measurement data shown in FIGS. 8A to 8C are binarized so that “1” is on the lead and “0” is between the leads. Foreign matter detection is "1"
The number of consecutive pixels of “0” is compared with the performance value and the lower limit value of the lead width determination set for each area, and the number of consecutive pixels of “0” is compared with the upper limit value and the lower limit value of the inter-read determination value set for each area. However, if there is a continuous number outside the range of the determination value, it is a method of determining that foreign matter is attached.

For example, the images of the foreign matters 11 and 12 shown in FIG. 2 are converted into the foreign matter 1 shown in FIG.
If the binarized images 1 and 12 are obtained, the graphs of the projection measurement data shown in FIG. 8 are deformed as shown by the waveforms 13 and 14, and the waveform 15 of the graph of FIG. ,
As shown in 16, the number of consecutive pixels in the "1" portion becomes abnormally small or large. Therefore, these consecutive pixel numbers are compared with the upper limit value and the lower limit value, and it is detected that they are out of the predetermined range. Then, it is determined that foreign matter is attached. The width of the flat portion 23 of the lead is narrower than the width of the shoulder portion 25, but the upper and lower limits of the lead width determination are set to the flat portion image 26, the slope portion image 27, or the shoulder portion image 28 for each area. By setting to, it is possible to determine that a foreign matter smaller than the lead width of the shoulder 25 adheres to the flat portion 23 as a foreign matter.

[0018]

According to the IC foreign matter inspection apparatus and method of the present invention, the IC lead portion has a shoulder portion having a high reflectance and a small lead width, a slope portion having a low reflectance, and a flat portion having a high reflectance and a large lead width. Since the lead foreign matter inspection is carried out for each of the three areas, the binarization level, lead width determination value, and lead-to-lead determination value suitable for each area can be set. This has the effect of improving the performance of detecting foreign matter on the leads with a small difference in gray value or foreign matter adhering between the leads.

[Brief description of drawings]

FIG. 1 is a block diagram of an IC foreign matter inspection apparatus according to an embodiment of the present invention.

FIG. 2 is a pattern diagram showing stored data d shown in FIG.

FIG. 3 is a pattern diagram of the binarized image signal e shown in FIG.

FIG. 4 is a diagram of a graph of measurement data f shown in FIG.

5 is a diagram of a graph of the projection binarized data g shown in FIG.

6 is a diagram of an image obtained by dividing the storage data d shown in FIG. 1 by a read area division signal h.

7 is a diagram of an image obtained by binarizing the divided image shown in FIG. 6 by the divided area foreign matter detection unit 10 shown in FIG.

FIG. 8 is a diagram of data obtained by projecting the divided binarized image shown in FIG. 7 in the lead longitudinal direction.

9 is a diagram of binarized data of the projected data shown in FIG. 8. FIG.

FIG. 10 is a block diagram showing a conventional IC foreign matter inspection method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 camera 2 IC 3 AD conversion means 4 inspection area cutting means 5 image storage means 6 first binarization means 7 X projection means 8 second binarization means 9 lead area division means 10 foreign matter detection means for each divided area 21 Lead 22 Mold 23 Flat part 24 Slope part 25 Shoulder part 26 Flat part region 23 'Flat part region 24' Slope part region 25 'Shoulder region 26 Flat part image 27 Slope part image 28 Shoulder image 31 Image sensor 32 Frame memory 33 Arithmetic processing unit a Analog image data b Grayscale image data c Inspection area grayscale image data d Storage data e Binarized image data f Measurement data g Projected binarized data h Lead area division signal

Claims (4)

[Claims] 1. A camera for taking an image of an IC in which a lead to be inspected is composed of a shoulder portion of a flat portion on the mold side, a flat portion on the tip side, and a slope portion located between these, and image data from the camera. AD conversion means for inputting and performing AD conversion to output grayscale image data; inspection area cutout means for cutting out the inspection area grayscale image data in a range including all leads for one side of the IC from the grayscale image data; First binarizing means for binarizing the area grayscale image data so that only the flat portion and the shoulder portion of the lead having a large amount of reflected light received by the camera become "1"; and the first binarizing Projection means for inputting the binarized image data binarized by the means and outputting X projection data obtained by measuring the number of pixels of "1" in each image element array in the X direction perpendicular to the longitudinal direction of the lead;
A portion of the inspection area grayscale image data where the X projection data is smaller than a predetermined value and which corresponds to the central portion of the lead is defined as a slope image and is larger than the predetermined value.
A lead area dividing unit that makes a portion corresponding to a mold side portion of C a shoulder image and a portion that is larger than the predetermined value and that corresponds to a tip side portion of the lead as a flat portion image, the shoulder image, Foreign object detection for each divided area in which the slope image and the flat image are binarized with a unique binarization level to identify the "1" pixel as the lead portion and detect foreign substances on the lead A foreign matter inspection device for an IC, comprising: 2. The foreign matter detection means for each divided area is a shoulder image,
The projection data obtained by measuring the number of pixels of "1" in each pixel row in the direction parallel to the longitudinal direction of the lead for the binarized image of each of the slope image and the flat image is binarized data. 2. The IC foreign matter inspection apparatus according to claim 1, wherein it is detected whether or not the width of the "1" portion or the "0" portion is within a range delimited by a predetermined upper limit value or lower limit value. 3. An IC having a shoulder of a flat portion on the mold side, a flat portion on the tip side, and a slope portion located in the middle between the leads to be inspected is picked up by a camera, and from the grayscale image data AD-converted. The inspection area grayscale image data in a range including all the leads for one side of the IC is cut out, and the amount of reflected light received by the camera for the inspection area grayscale image data is large, and only the flat portion and the shoulder portion of the lead are "1". X of the binarized data binarized so that the number of "1" pixels in each pixel row in the X direction perpendicular to the longitudinal direction of the lead is measured.
Projection data is divided by a predetermined value, the inspection area grayscale image data is divided, and the slope image of a portion corresponding to the center of the lead in a portion where the X projection data is smaller than the predetermined value, the predetermined value A shoulder image of a larger portion corresponding to the mold side portion and a flat portion image of a portion greater than the predetermined value and corresponding to the tip end side portion of the lead are each binarized at a unique binarization level. A method of inspecting a foreign matter for an IC, which is characterized by identifying the lead portion and detecting a foreign matter attached to the lead. 4. A binarized image obtained by binarizing a slope image, a shoulder image, and a flat image with a unique binarization level for each pixel row in the direction parallel to the longitudinal direction of the lead.
The projection data obtained by measuring the number of pixels of "1" is obtained, and the width of the "1" part or the "0" part of the binarized data of this projection data falls within a range delimited by a predetermined upper limit value or lower limit value. 4. The IC foreign matter inspection method according to claim 3, wherein the foreign matter attached to the lead is detected depending on whether or not the lead is present.