JP2014095661A - Center position detection method and center position detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable any operator to obtain the same result of detection of the center position of a region to be detected as long as the detection is carried out under the same conditions and also to complete the detection in a short time.SOLUTION: A center position detection method is designed to detect the position of the center of a region to be detected from a specified region specified for image data, and includes a binarization step S5, a retrieval step S6, and a rectangle operation step S7 which are automatically executed. The binarization step S5 comprises binarizing the specified region so as to obtain a binarized image. The retrieval step S6 comprises retrieving the widest region in the binarized image obtained through the binarization step S5. The rectangle operation step S7 comprises computing the largest rectangle and square inscribed in the widest region retrieved through the retrieval step S6. The rectangle operation step S7 comprises determining, as the center position of the region to be detected, the center position of the largest inscribed rectangle and square computed through the rectangle operation step S7.

Description

  The present invention relates to a center position detecting method and a center position detecting apparatus used for detecting a center position in an island to which a chip of a semiconductor wafer is attached, for example.

  In a manufacturing process of a semiconductor device such as an LSI, a plurality of chips are formed on the surface of a semiconductor wafer and then separated into chips by dicing. Then, of these separated chips, those which are determined to be non-defective by electrical inspection before dicing are picked up and bonded to a lead frame or the like.

  Among the manufacturing processes, in the bonding process, the picked-up chip is bonded to a portion called an island such as a lead frame with silver paste or the like. At this time, it is necessary to match the center position of the chip with the center position of the island. For this reason, the center position of the chip and the center position of the island are automatically calculated (detected) from the image data (see, for example, Patent Document 1).

JP-A-10-313013

  By the way, unlike the situation where the center position of the island must be detected continuously during the bonding process during the manufacture of semiconductor devices, the center position of the island must be detected in situations where the center position of the island is detected only a few times. In some cases. This is because in order to perform the automatic detection as described above, an island model must be registered in advance, and the work for this is complicated.

  Detection of the center position of the island by the operator is performed as follows while the operator looks at the screen of a monitor or the like. For example, as shown in FIG. 7, when the image of the island I ′ of the lead frame L ′ is displayed on the screen of a monitor or the like, the center of the cursor C ′ indicated by the dotted line (cross intersection) is set to the island I ′. Set to the center position of.

  However, since the detection by the operator is a visual alignment operation, an alignment error occurs due to the operator's sense, and it takes time for alignment. This is due to the following reason. Even if the island has an accurate point-symmetric shape (for example, a rectangular shape) by design, it is rare to have an accurate point-symmetric shape due to an error in manufacturing, and it is difficult to align. Further, since the image itself has a resolution, an error of one pixel is likely to occur by an operator. For this reason, when the accuracy is required, the alignment position needs to be corrected many times.

  Such a problem occurs not only in the center position of an island such as a lead frame, but also when the center position of a detected part in other members is detected by an operator.

  In view of the above circumstances, it is an object of the present invention to obtain the same result and to finish the detection in a short time for any worker with the same conditions for the detection of the center position of the detected part.

  The present invention for solving the above-described problem is a method for detecting the position of the center of a detected region from a specified region specified to include the detected region in the image data, and binarizing the specified region A binarization process for obtaining a binarized image, a search process for searching for the widest area among the binarized images obtained in the binarization process, and the widest area searched for in the search process, A calculation step of calculating a maximum inscribed rectangle or square, and the center position of the maximum inscribed rectangle or square calculated in the calculation step is set as the center position of the detected portion. It is.

  In this method, the widest area searched in the search process corresponds to the detected part. Since the center position of the maximum inscribed rectangle or the maximum inscribed square calculated in the calculation step corresponds to the center position of the detected part, this is set as the center position of the detected part. Therefore, the center position of the detected part can be detected from the designated region by the binarization process, the search process, and the calculation process.

  The information input by the operator is only the designated area including the detected part, and parameters and the like can be made unnecessary. And after designating a designation | designated area | region, the binarization process, a search process, and a calculation process can be performed automatically, and the center position of a to-be-detected site | part can be detected.

  If the binarization process, search process, and calculation process are performed automatically, even if the designated area is manually specified for the detection of the center position of the detected part, a certain result is obtained and the time required for detection can be shortened. .

  In addition, when an operator designates a designated area, it is only necessary to include a detected part, so that accuracy is not required and the time required for area designation is short. Further, since the worker does not need to input information other than designating the designated area, the working time can be suppressed.

  That is, in the center position detection method of the present invention, the same result can be obtained for any worker under the same conditions for the detection of the center position of the detected part, and the detection can be completed in a short time.

  Further, the center position of the maximum inscribed rectangle or square with respect to the widest area corresponding to the detected part is used as the central position of the detected part. If the maximum inscribed rectangle or square is used in this way, it is possible to cope with a certain amount of shape error that occurs during the manufacture of the detected part. In addition, when the maximum inscribed rectangle or square is used, it is possible to cope with a detected portion having a shape other than the rectangular shape to some extent.

  In the above method, the average luminance of the designated area is calculated between the designation of the designated area and the binarization process, and a threshold value for obtaining a binarized image in the binarization process is calculated based on the calculated average luminance. You may provide the threshold value setting process to set.

  If the threshold set in this threshold setting step is used to obtain a binarized image in the binarization step, an appropriate binarized image can be obtained. Moreover, this threshold value setting process can also be automated, and an increase in the overall detection time can be suppressed.

  Furthermore, if a smoothing step for smoothing the designated region is provided between the designation of the designated region and the threshold setting step, noise can be reduced by smoothing, and a more accurate detection result can be obtained. Moreover, this smoothing process can also be automated and it can suppress that the whole detection time increases. Note that smoothing refers to a process of removing noise generated in an image or reducing a change in luminance value for each pixel by using a moving average filter or the like.

  In any of the above methods, in the search step, a region having an area of 40 to 90% with respect to the area of the designated region may be searched as the widest region.

  This search can simplify the algorithm when automated, thereby shortening the processing time. Therefore, the time required for detecting the center position can be further shortened.

  In any of the above methods, the detected site may be an island to which a semiconductor wafer chip is attached.

  In this case, the above-described effects can be enjoyed as being particularly effective.

  According to another aspect of the present invention, there is provided an apparatus for detecting the position of the center of a detected part from a specified region specified to include the detected part with respect to the image data, and acquiring the image data. Image acquisition means for processing and image processing means for processing image data, the image processing means binarizing means for binarizing a designated area to obtain a binarized image, and this binarizing means Search means for searching for the widest area among the binarized images obtained in (2), and an arithmetic means for calculating the largest rectangle or square inscribed in the widest area searched by the search means, The center position detecting device is characterized in that the center position of the maximum inscribed rectangle or square calculated by the calculating means is set as the center position of the detected part.

  In this apparatus, since the method described at the beginning can be performed, the same effect as this method can be obtained.

  According to the present invention, regarding the detection of the center position of the detected part, any worker can obtain the same result under the same conditions, and the detection can be completed in a short time.

It is a schematic diagram which shows the center position detection apparatus which concerns on embodiment of this invention. It is a figure which shows the flowchart of the center position detection method which concerns on embodiment of this invention. It is a figure which shows the screen by the center position detection method which concerns on embodiment of this invention. It is a figure which shows the screen by the center position detection method which concerns on embodiment of this invention. In the schematic diagram which shows a to-be-detected site | part, (A) is a normal state, (B) is a state with a large shape error. It is a schematic diagram which shows the other example of a to-be-detected site | part, (A) is an elliptical shape, (B) is a perfect circle shape, (C) is a square shape. It is a figure which shows the screen by the conventional center position detection method.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings.

  FIG. 1 schematically shows a center position detection apparatus according to an embodiment of the present invention. In the present embodiment, the center position detection device 1 detects the center position of an island as a detected part for image data of the lead frame L.

  The center position detection device 1 is operated by an operator by an image acquisition means 2 such as a camera for acquiring image data of the lead frame L, an image processing means 3 such as a personal computer, an image display means 4 such as a monitor, and the like. The designated area 5 such as a mouse is a main component.

  The image processing means 3 includes an image control means 3a, a smoothing means 3b, a threshold setting means 3c, a binarization means 3d, a search means 3e, a rectangle calculation means 3f, and a center position calculation means 3g. Storage control means 3h and storage means 3i.

  The image control unit 3a controls an image to be displayed on the image display unit 4. The smoothing means 3b smoothes the area specified by the operator for the acquired image data. The threshold setting unit 3c calculates the average luminance of the designated area smoothed by the smoothing unit 3b, and sets a threshold for obtaining a binarized image by the binarizing unit 3d based on the calculated average luminance. . The binarizing means 3d binarizes the designated area smoothed by the smoothing means 3b with the threshold set by the threshold setting means 3c to obtain a binarized image.

  The search means 3e divides the binarized image obtained by the binarization means 3d, and searches the widest area among them. The rectangle calculation means 3f calculates the maximum rectangle inscribed in the widest area searched by the search means 3e by calculation. Then, the center position calculation means 3g calculates the center position of the maximum inscribed rectangle obtained by the rectangle calculation means 3f by calculation. The storage control means 3h causes the storage means 3i to store the center position of the maximum inscribed rectangle obtained by the center position calculation means 3g as the island center position.

  The means included in these image processing means 3 can be constituted by, for example, software (program) in a personal computer, a memory, or the like.

  Next, a center position detection method used in the center position detection apparatus 1 will be described.

  As shown in FIG. 2, the center position detection method of the present embodiment includes an image display step S1, a region designation step S2, a smoothing step S3, a threshold setting step S4, a binarization step S5, and a search step. S6, a rectangle calculation step S7, a center position calculation step S8, and a center position processing step S9, which are executed in the order indicated by the arrows in FIG.

  In the image display step S1, the image data of the lead frame L acquired by the image acquisition unit 2 is displayed on the image display unit 4 by the image control unit 3a as shown in FIG. In this embodiment, the island I of the lead frame L has a rectangular shape with slightly rounded corners, and is raised by pressing (on the front side of the paper surface in FIG. 3).

  Next, in the area designating step S2, the operator designates an area where the island I exists for the image displayed on the image display means 4 in the image display process S1. Specifically, the operator operates the area specifying means 5 to move the cursor C1 including a cross indicated by a one-dot chain line and a square indicated by a dotted line in FIG. Then, when the square frame (dotted line portion) of the cursor C1 is in a state including the island I, an operation for determining the position is performed. As a result, the area surrounded by the square of the cursor C1 becomes the designated area R designated by the operator. At this time, the position of the cross (dashed line) indicating the center of the cursor C1 may not be the center of the island I. The display of the cursor C1 on the image display means 4 is performed by the image control means 3a.

  When the operator performs an operation of determining the position of the cursor C1 in the area designating step S2 (confirming designation of the designated area R), the smoothing process S3 to the center position processing process S9 are automatically performed in the following order. To be executed.

  In the smoothing step S3, the smoothing means 3b smoothes the image data of the designated region R, and noise is removed by this smoothing.

  In the threshold setting step S4, the threshold setting means 3c calculates the average luminance of the image data of the designated region R smoothed in the smoothing step S3, and binarizes in the binarization step S5 based on the calculated average luminance. Sets the threshold used to obtain the image.

  In the binarization step S5, the binarization means 3d binarizes the smoothed image data of the designated region R using the threshold set in the threshold setting step S4 to obtain binarized image data.

  In the search step S6, the search means 3e divides the designated region R as the binarized image data obtained in the binarization step S5 into a plurality of regions, and among these divided regions, as shown in FIG. The widest area D is searched. The searched widest region D corresponds to the island I. In the present embodiment, the search unit 3e searches for a region having an area of 40 to 90% with respect to the area of the designated region R as the widest region D. That is, the search means 3e removes a region having an area of less than 40% or more than 90% of the area of the designated region R.

  In the rectangle calculation step S7, the rectangle calculation means 3f calculates and calculates the maximum inscribed rectangle P as shown by the dotted line in FIG. 5A for the widest area D searched in the search step S6.

  In the center position calculation step S8, the center position calculation means 3g calculates and determines the center position O of the maximum inscribed rectangle obtained in the rectangle calculation step S7 as shown in FIG. The center position O of the maximum inscribed rectangle P corresponds to the center position of the island I.

  In the center position processing step S9, the image control means 3a sets the center position O of the maximum inscribed rectangle P obtained in the center position calculation step S8 as the center position of the island I, as shown in FIG. The image is displayed on the display means 4 with a cross cursor C2. In the center position processing step S9, the storage control unit 3h stores the center position O of the maximum inscribed rectangle P obtained in the center position calculation step S8 in the storage unit 3i as the center position of the island I.

  Then, the above operation is repeated, and the center positions of the plurality of islands I are stored in the storage unit 3i for one lead frame L. The stored center positions of the plurality of islands I are used to calculate the center positions of other islands I as follows, for example. Consider a case where the lead frame L is rectangular and the islands I are arranged in a matrix at equal pitches along the long side direction and the short side direction of the lead frame L. In this case, the entire matrix is rectangular and there are four corners. The center position of the island I arranged at three of these corners is stored in the storage means 3i. Then, all the center positions of the other islands I can be calculated from the pitch interval and the stored center positions of the three islands I.

  In the center position detection apparatus 1 and the center position detection method of the present embodiment described above, the information input by the operator is only the designated region R including the island I and does not require parameters or the like. Then, after designating (determining) the designated region R, software or the like automatically executes the smoothing step S3 to the center position processing step S9 to detect the center position of the island I.

  Since the smoothing step S3 to the center position processing step S9 are automatically performed, even if the operator manually designates the designated region R with respect to the detection of the center position of the island I, a constant result is obtained. The time required can be shortened.

  Further, when the operator designates the designated region R on the screen of the image display means 4, it is only necessary to include the island I. Therefore, accuracy is not required and the time required for designating the designated region R is short. It becomes. Further, since the worker does not need to input information such as parameters other than designating the designated region R, the work time can be suppressed.

  That is, in the center position detection apparatus 1 and the center position detection method of the present embodiment, the same result can be obtained for any worker under the same conditions for the detection of the center position of the island I, and the detection can be performed in a short time. Can finish.

  Further, as the center position of the island I, the center position O of the maximum inscribed rectangle P with respect to the widest region D corresponding to the island I is used. If the maximum inscribed rectangle P is used as described above, it is possible to cope with a shape error caused when the island I is manufactured to some extent. For example, as shown in FIG. 5B, even when the corner of the island I is rounded due to a shape error when the lead frame L is manufactured, the maximum inscribed center position O is substantially the same as in the case of FIG. A rectangle P is obtained. Therefore, it is possible to suppress the deterioration of the detection accuracy of the center position of the island I due to the shape error.

  Further, when the maximum inscribed rectangle P is used, it is possible to cope with the island I having a shape other than the rectangular shape to some extent. For example, for the elliptical island I shown in FIG. 6A, the center position O of the maximum inscribed rectangle P is the center position of the island I. 6B, the center position O of the maximum inscribed rectangle P is also the center position of the island I. As shown in FIG. For the perfect circular island I shown in FIG. 6B or the square island I shown in FIG. 6C, the maximum inscribed square P ′ is obtained by calculation, and the center position thereof is obtained. O can be the center position of the island I.

  In addition, since the image data of the designated region R is smoothed in the smoothing step S3, noise can be reduced by smoothing, and a more accurate detection result can be obtained.

  Further, as a threshold for obtaining a binarized image in the binarization step S5, a threshold based on the average luminance calculated from the image data of the designated region R smoothed in the smoothing step S3 is used. Therefore, an appropriate binarized image can be obtained in the binarization step S5.

  In the search step S6, a region having an area of 40 to 90% with respect to the area of the designated region R is searched as the widest region. This simplifies the algorithm, thereby reducing the time required for the search step S6. Therefore, the time required for detecting the center position can be further shortened.

  The present invention is not limited to the above-described embodiment, and various modifications can be made within the scope of the technical idea. For example, in the above-described embodiment, the center of the island of the lead frame is detected, but the center of the detection site in a member other than the lead frame may be detected. Moreover, in the said embodiment, although the area | region designation | designated process was performed manually by the operator, you may automate.

DESCRIPTION OF SYMBOLS 1 Center position detection apparatus 3b Smoothing means 3c Threshold setting means 3d Binarization means 3e Search means 3f Rectangle calculation means D The widest area | region I Island (detection site | part)
O center position P maximum inscribed rectangle P ′ maximum inscribed square R designated area S3 smoothing step S4 threshold setting step S5 binarization step S6 search step S7 rectangle calculation step

Claims (6)

A method for detecting the position of the center of a detected part from a designated region specified to include the detected part for image data,
A binarization step of binarizing a specified area to obtain a binarized image;
Among the binarized images obtained in this binarization step, a search step for searching the widest region,
A calculation step for calculating the largest rectangle or square inscribed for the widest area searched in the search step,
A center position detection method characterized in that the center position of the maximum inscribed rectangle or square calculated in this calculation step is set as the center position of the detected part.   A threshold setting step for calculating the average luminance of the designated region between the designation of the designated region and the binarization step, and setting a threshold for obtaining a binarized image in the binarization step based on the calculated average luminance The center position detection method according to claim 1, further comprising:   The center position detection method according to claim 2, further comprising a smoothing step of smoothing the designated region between the designation of the designated region and the threshold value setting step.   The center position detection method according to any one of claims 1 to 3, wherein an area having an area of 40 to 90% with respect to an area of the designated area is searched as a widest area in the search step.   The center position detection method according to claim 1, wherein the detected part is an island to which a chip of a semiconductor wafer is attached. A device that detects the position of the center of a detected part from a specified region that is specified to include the detected part for image data,
Image acquisition means for acquiring image data, and image processing means for processing the image data,
The image processing means includes a binarizing means for binarizing the designated area to obtain a binarized image, a search means for searching for the widest area among the binarized images obtained by the binarizing means, An operation means for calculating the largest rectangle or square inscribed for the widest area searched by the search means;
A center position detecting apparatus characterized in that a center position of a maximum inscribed rectangle or square calculated by the calculating means is set as a center position of a detected part.

Images