JP2004077284A - Locating method of object having recursive structure - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve positioning precision more than the conventional method by performing calculation from design data of a chip. <P>SOLUTION: The locating method of an object having recursive structure obtaining offset of position of a wafer, searches a recursive part from an image, and determines to which recursive part of a chip it corresponds from arrangement of the recursive part. From the chip in the image, a direction where a reference position of the chip is determined, and coordinates of the reference position of a chip is obtained. Based on an angle of a wafer relative to the horizontal direction, the wafer is rotated and set horizontally. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a method for measuring the position of an object having a repetitive structure using image search and performing alignment (positioning) on the position of an object having a repetitive structure. Here, the repetitive structure is a structure in which terminals such as balls and pads of the same shape are arranged in a fixed direction, and the alignment means performing position measurement (chip position, angle) of a work.
[0002]
[Prior art]
In order to measure the characteristics of a semiconductor mounted on a substrate such as a wafer level CSP (Chip Size Package) or a package package, an electrode must be brought into contact with a ball or a pad which is a terminal of each semiconductor element. Therefore, the position of the semiconductor must be accurately measured.
[0003]
FIG. 6 is a diagram schematically showing a circuit composed of a plurality of semiconductor elements. For example, FIG. 6A shows a work packaged in a collective mold, in which a plurality of devices 51 are arranged.
[0004]
FIG. 6B is an enlarged view of one device. In this example, a plurality of balls 511 as terminals are regularly formed in each device. An electrode is pressed against the ball to measure the characteristics of the device.
[0005]
In order to accurately press the electrode against the ball 511, the position of the ball 511, that is, the position of the device must be accurately known. For this purpose, at the time of teaching, one device 51 or an arrangement of characteristic balls 511 in the device 51 is registered as a reference image.
[0006]
The reference image is an image obtained by photographing a reference device. In addition, in order to calculate the reference position of the device, the offset relationship between the reference image and the reference point on the device is also registered at the time of teaching.
[0007]
Next, the work 50 to be measured is photographed, and a search is made as to whether or not there is a reference image registered in the photographed image. When an image that matches the reference image is found, the reference position of the device is calculated from the offset relationship between the registered reference image and the reference point on the device.
[0008]
When this operation is repeated with some devices on the work 50 to be measured, for example, devices at four corners, the position of the work can be calculated. This position measuring method uses a semiconductor wafer position measuring technique.
[0009]
[Problems to be solved by the invention]
However, since a wafer has a circuit pattern, there is a region having a unique pattern in one chip. Therefore, even in a large chip, it is possible to search for a place where the same pattern exists, using a small area as a reference. However, in the wafer level package, the only clue for alignment is the terminal arrangement on the terminal surface. If a region where the terminal arrangement has a unique pattern comes into view, alignment can be performed, but such a unique terminal arrangement does not always exist.
[0010]
Even if there is a region having a unique terminal arrangement, it is necessary to inform the device of the region in order to process the type. The accuracy in teaching this area directly affects the actual positioning. And the error at the time of teaching becomes the systematic error of positioning.
[0011]
Further, in the conventional method of using the image of the chip used in the prober as the reference image, it is possible to determine whether or not the reference image exists in the image. Therefore, if it is in the image, the position can be determined. If not, it is inefficient because you have to search the surroundings in order.
[0012]
The present invention has been made in view of the above circumstances, instead of teaching the chip area and the like, by calculating from the design data of the chip, to increase the positioning accuracy than the conventional method, further reduce the teaching work, An object of the present invention is to provide a method for measuring the position of an object having a repetitive structure that realizes alignment without depending on the size of a chip.
[0013]
[Means for Solving the Problems]
In order to achieve the above object, an invention according to claim 1 is a method for measuring the position of an object having a repeating structure, wherein the position of the object having the repeating structure is measured using image retrieval, wherein Searching from the top, judging which repetition part of the chip corresponds to the arrangement of the repetition part, judging the direction where the reference position of the chip is from the chip in the image, obtaining the coordinates of the reference position of the chip, The method is characterized in that the object is rotated to be horizontal based on the angle of the object with respect to the horizontal, and an offset of the position of the object is obtained.
[0014]
A method for measuring the position of an object having a repeating structure, wherein the position of the object having the repeating structure is measured using an image search, wherein an image of a repeating portion of the object having the repeating structure is generated from design data; Searching for an image of a repetitive portion of the object having the repetitive structure as a reference image, and determining which repetitive portion of the chip of the object having the repetitive structure corresponds to the arrangement of the repetitive portion of the searched image. Determining; determining a direction in which a reference position of the chip is located based on the number of the chips in a previous image; moving the image; and obtaining coordinates of the reference position of the chip; Find the angle of the object with the horizontal with respect to the horizontal, rotate the object having the repeating structure, and have the repeating structure A step to level the object, characterized in that it comprises the the steps of calculating an offset of an object having the repeating structure from the design data based on the angle of rotating the object having the repeating structure.
[0015]
Therefore, according to the first and second aspects of the present invention, the object to be searched in the image is a contact (ball) which is a repetitive portion, the contact is searched from the image, and the arrangement corresponds to any contact of the chip. And determine the direction in which the reference position of the chip is located based on the number of chips in the image, move the coordinates of the reference position of the chip, determine the angle of the object having the repetitive structure with respect to the horizontal, and determine the angle by the angle. Rotation is made horizontal, and based on the rotation, an offset of the position of the object having the repetitive structure can be obtained from the design value.
[0016]
According to a third aspect of the present invention, in addition to the first and second aspects, the reference position of the chip is any one of the four corners of the chip.
[0017]
Therefore, according to the third aspect of the present invention, since the reference position of the chip is any of the repeated portions (balls) at the four corners of the chip, the offset calculation is performed based on the interval between the repeated portions (ball interval). It can be easier.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0019]
FIG. 1 shows an outline of an apparatus for implementing the method for measuring the position of an object having a repeating structure according to the present invention. The work 1 is placed on the drive unit 2, and a camera 3 is provided above the work 1 so as to capture an image of the work.
[0020]
The camera 3 is connected to an image capturing unit 4 that captures an image from the camera 3. The image capturing unit 4 is connected to an arithmetic control unit 5 that processes an image captured by the image capturing unit 4. The arithmetic control unit 5 is connected to a motor control unit 6 that moves the position of the work based on a command from the arithmetic control unit 5, and the motor control unit 6 receives a signal from the motor control unit 6, It is connected to a drive unit 2 for positioning a target work 1.
[0021]
The drive unit 2 can drive four axes of movement in the XY directions, height movement in the Z direction, and θ for rotating the work 1. The XY axes move the work 1 in the horizontal direction. The Z axis adjusts the work 1 to the focal plane height of the camera lens. The θ axis rotates the work 1 so as to be parallel to the XY axes.
[0022]
In the method for measuring the position of an object having a repeating structure according to the present invention, first, the position of the upper left corner of the chip near the center is determined. From there, it moves horizontally by several chips to find the chip position. The inclination of the work 1 is corrected from the angle between the two points. Since it is almost horizontal, it moves to the chip near the outermost periphery and performs angle correction again. From this result, a position offset from when the work 1 is at a predetermined position can be calculated. FIG. 2 shows an overall schematic flow, and FIG. 3 shows a positional relationship of chips for obtaining positions. The steps in the general flow are the same as those in the conventional prober alignment.
[0023]
The method for measuring the position of an object having a repeating structure according to the present invention will be described in detail with reference to FIGS.
[0024]
First, the processing of the first point is performed as follows. The first point is that the chip at the center of the work is selected based on the calculation of the position of the upper left corner of the chip at the center of the work 1 (step S1 in FIG. 2). If the chip size is small, multiple chips appear in the field of view. If it is left as it is, it will be confused, so the area of the chip size + α is set as the work target area. If the chip is large, the entire image becomes the work area. The following description is of processing in this work area.
[0025]
The image is fetched, and the contact portion (ball) in the image is searched for using a search method such as a normalized correlation method, using the shape of the contact portion defined geometrically as a reference image. Classify by chip based on ball spacing.
[0026]
The work area can be classified into one, two, and four chips. In the case of one chip (FIG. 4- (1)), since there is a space at the upper left that is equal to or greater than the ball interval in the chip, the position of the upper left corner of the chip can be specified. In the case of FIG. 4- (2), since there are balls at the ball interval in the chip, the ball moves so as to show the upper left direction where the upper left exists (step S2 in FIG. 2). In the case of FIG. 4- (3), since two chips are shown at the top and bottom, it moves so that the top is shown so as to be in the state of FIG. 4- (2) (step S2 in FIG. 2). In the case of FIG. 4- (4), since two chips are shown on the left and right, the camera is moved so that the left is shown so as to be in the state of FIG. 4- (2) (step S2 in FIG. 2). In the case of FIG. 4- (5), since four chips are photographed, the image is moved so that the upper left is photographed so as to be in the state of FIG. 4- (2) (step S2 in FIG. 2).
[0027]
The movement is repeated to make the image appear as shown in FIG. Then, an image captured as shown in FIG. 4- (1) is captured from the camera 3 (step S3 in FIG. 2), and the image is processed to search for a corner (step S4 in FIG. 2). As a result of the search, it is determined whether an image shown in FIG. 4- (1) is found (step S5 in FIG. 2), and if not found, the work 1 is moved and the upper left corner is searched (FIG. 4). Step S6 in Step 2). Again, it is determined whether or not an image shown in FIG. 4- (1) is found (step S7 in FIG. 2), and if not found, an error is processed (step S8 in FIG. 2).
[0028]
When FIG. 4- (1) is reached (FIG. 4- (1) is found), the position of the ball at the upper left corner can be determined. Here, the upper left corner is set as a target, but when a plurality of chips are photographed, other corners can be moved as targets in consideration of the number of balls in the chips.
[0029]
The processing of the next second point is performed as follows. First, in order to make the work 1 roughly horizontal, search for a chip that is slightly away. The next chip to be searched is calculated so that the chip in the same row as the chip size always enters the image (step S9 in FIG. 2). The position of the second point is a distance at which the variation due to the transfer as shown in FIG. 5 falls within the image (field of view). Within this range, there is no possibility that the position of the chip in another row is obtained. The search for the second chip is performed in the same manner as the first chip shown in FIG.
[0030]
Specifically, an image of the second chip taken from the camera as shown in FIG. 4- (1) is captured (step S10 in FIG. 2), and the image is processed to search for the same corner (FIG. 2). Step S11). If the data cannot be captured, the workpiece is moved and searched by the same processing as the first point. Then, as a result of the search, it is determined whether or not the same corner is found (step S12 in FIG. 2), and if not found, it is processed as an error (step S13 in FIG. 2). Correct the obtained angle and level the work. That is, the angle between the first point and the two-point plane is calculated, and the work 1 is rotated by θ so as to be horizontal (step S14 in FIG. 2).
[0031]
The processing near the last outermost periphery is performed as follows. To improve the accuracy of the angle, it is necessary to measure the angle between points as far apart as possible. Specifically, it moves horizontally and moves so that the same corner of the outermost chip is in the camera view (step S15 in FIG. 2). The chip position near the outermost periphery of the work 1 is obtained, and the chip position is obtained. At this time, the same search method as the first point may be used. Also, the corners are to be found at the first position, because it is somewhat horizontal.
[0032]
In order to find a corner, similarly to the search for the first and second points, an image is fetched from the camera (step S16 in FIG. 2), and the image is processed to search for the same corner (step in FIG. 2). S17). As a result of the search, it is determined whether or not the same corner is found (step S18 in FIG. 2), and if not found, it is processed as an error (step S19 in FIG. 2). If found, the work is rotated by θ so as to be horizontal (step S20 in FIG. 2). Then, an offset between the calculated value of the chip position and the actually measured value is obtained (step S21 in FIG. 2).
[0033]
As another embodiment of the present invention, similarly to a wafer, a characteristic partial image of an actual chip can be used as a reference image. In this case, since there is little information on the field of view currently being viewed, the surroundings of the first field of view are searched sequentially for partial images identical to the reference image. Therefore, the processing efficiency decreases.
[0034]
In addition, here, the camera is constituted by one unit, but the positioning accuracy can be improved by adding a camera having a high-magnification lens. In this case, after the flow of FIG. 2, a process of measuring the chip position again with the high-magnification camera and calculating the angle and the offset is added.
[0035]
The present invention can be applied not only to a CSP at a wafer level but also to a strip (batch molding) having a rectangular outer shape.
[0036]
【The invention's effect】
According to the present invention, since chip information in an image can be used, a chip position can be efficiently searched. In particular, since the method does not use the image of the chip as the reference image as in the related art, even when the reference image does not exist in the image, it is not necessary to search the periphery in order, and the work efficiency can be improved.
[0037]
Further, according to the present invention, the reference image of the ball is geometrically generated from the design information of the chip, so that it is not necessary to teach the image. In particular, there is no need for the experience of what part of the chip to teach the actual image to the system when it is taught by the conventional method.
[0038]
Further, there is a case where the reference image and the position of the chip are associated with one of the teaching purposes. The user specifies a corner of the chip, and calculates the offset of the origin of the reference image from the corner. If the position of the corner of the chip taught by the user is displaced, it is directly connected to the position displacement of the work. In the present invention, since the relationship between the center position of the ball and the position of the chip is known as a design value, no error occurs from here.
[Brief description of the drawings]
FIG. 1 is a schematic diagram of an apparatus for performing a position measuring method of an object having a repeating structure according to the present invention.
FIG. 2 is a flowchart showing a processing flow of a method for measuring the position of an object having a repeating structure according to the present invention.
FIG. 3 is an explanatory diagram showing a positional relationship of chips whose position is determined by alignment.
FIG. 4 is an explanatory diagram showing a relationship between a chip shown as an image and a moving direction.
FIG. 5 is an explanatory diagram showing how to determine the position of a second point.
FIG. 6 is an explanatory diagram for explaining a conventional position measuring method.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Work 2 Drive part 3 Camera 4 Image taking-in part 5 Operation control part 6 Motor control part

Claims (3)

A method for measuring the position of an object having a repeating structure that measures the position of the object having a repeating structure using image search,
The repetition part is searched from the image, and from the arrangement of the repetition part, it is determined which repetition part of the chip corresponds to, the direction of the reference position of the chip from the chip in the image is determined, and the coordinates of the reference position of the chip are determined. And a method for measuring the position of the object having a repetitive structure in which the object is rotated to be horizontal based on the angle of the object with respect to the horizontal, and an offset of the position of the object is obtained. A method for measuring the position of an object having a repeating structure that measures the position of the object having a repeating structure using image search,
Generating an image of a repeating portion of the object having the repeating structure from the design data,
Searching for an image of a repeated portion of the object having the repeating structure using the image as a reference image,
A step of determining which repetitive part of the chip of the object having the repetitive structure corresponds to the arrangement of the repetitive part of the searched image,
Determining the direction in which the reference position of the chip is based on the number of the chips in the previous image, and moving the image;
Obtaining the coordinates of the reference position of the chip;
Determine the angle to the horizontal of the object having the repeating structure, rotate the object having the repeating structure, the step of making the object having the repeating structure horizontal, and based on the angle at which the object having the repeating structure is rotated, Calculating the offset of the object having the repeating structure from design data. A method for measuring the position of an object having a repeating structure. The position measuring method for an object having a repeating structure according to claim 1 or 2, wherein the reference position of the chip is any one of the repeated portions at four corners of the chip.

Images