JPH01173173A - Recognizing device containing position correcting function - Google Patents

Abstract

PURPOSE:To obtain the center between recognizing objects and to automatically lead a chip into a window frame by inverting the logic of the binary level of an image signal. CONSTITUTION:A converting part 15 converts the image signal received from a taking-in device into a black/white binary level. An image arithmetic processing part 16 gives an arithmetic operation to the converted digital signal and obtains the center between two recognizing objects from the pattern information received previously. A data inverting part 21 inverts selectively the logic of the binarized image signal inputted to th part 16. In case no chip is included in a window frame, the recognizing object is changed into a street from a chip with inversion of the logic of the binarized image signal inputted to the part 16. Then the position correction data is obtained based on the obtained data. Thus the chip is moved into the window frame with position correction and the position of the chip can be automatically corrected at the time of first position.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention is a recognition device that recognizes the center position, defective marks, cracks, etc. of a semiconductor chip (hereinafter abbreviated as a chip), etc., in which a recognition target is guided into a recognition range. The present invention relates to a recognition device equipped with a position correction function.

Conventional technology In recent years, semiconductor-related developments have been particularly remarkable, and memory I
As seen in CE, ultra-high integration is gaining momentum. On the other hand, there is also a lot of production of hybrid ICs that integrate dedicated circuits and combine them with other elements, and along with the bonding technology that attaches the chip to the substrate, it automatically recognizes the center position of the chip, whether it is good or bad, and whether it is cracked. Recognition technology is gaining importance. .

DESCRIPTION OF THE PREFERRED EMBODIMENTS A recognition device for a die bonding apparatus will be described below as a conventional example with reference to one drawing.

FIG. 4 shows a schematic configuration of the die bonding apparatus. In FIG. 4, 1 is an expanded wafer, and 2 is a wafer ring for fixing the expanded wafer. 3 is a recognition camera, 4 is a part of a half mirror for illumination, 6 is an optical fiber that guides light, and 6 is a light source for illumination. F is a recognition device that receives signals from the recognition camera 3 and performs recognition processing. 8 is a television for monitoring. 9 is an expanded chip. On the TV screen, 10.11 is netorito, the gap between chips 9 and 9. A window frame 12 indicates the recognition boundary line of this recognition device, and the inside of the rectangle is the recognition range. Reference numeral 13 is an x-Y table for feeding the chip 9. A head 14 picks up the chip 9, transfers it, and attaches it to the substrate 15.

The die bonding device attaches the chip 9 on the wafer ring 2 onto the substrate 16.The chip 9 is checked for electrical function in advance, and if it is defective, the chip 9 is removed. There is a red and black mark on the top to indicate that it is defective.

When the chip 9 is removed from the wafer ring 2, it is determined whether there are any defective chips and only good chips are mounted on the substrate 115. The recognition device 7 makes this judgment as to whether the product is good or bad.

In FIG. 6 showing the internal configuration of the Mfi device 7, 1
Reference numeral 6 denotes a binarization converter that converts the image signal (analog) into binary levels of white and black, and 16 an image arithmetic processing unit that performs arithmetic processing on the binarized image signal. 17 Hiro This is the monitor control section that outputs the image signal and the binarized signal to the monitor. 18 is an operation panel. Reference numeral 19 denotes an interface section 7i which receives signals from the operation panel 18 and an external controller. Reference numeral 2o denotes a central control section that controls the image calculation processing section 16 and the monitor control section 17 based on signals sent from the interface unit 9.

Which recognition device determines the area and center position of the object to be recognized and determines whether it is good or bad.

Send center position data.

The operation of the recognition device for the die bonding apparatus configured as described above will be explained below based on the schematic flowchart of the operation shown in FIG.

First, move the x-y table 13 laterally from the recognition position A as indicated by arrow a, replace the wafer ring 2, and set a new wafer on the x-y table 13 (
(not shown). Next, the recognition position A at the bottom of the recognition camera 3
Move the X-Y table 13 as shown by the arrow. After the positioning is completed, it is confirmed on the monitor television 8 whether the tip 9 has entered without touching the window frame 12. If window frame 12
Chip 9 is in contact with the ! in Figure 7! If the chip 9 is not inside the window frame 12 as shown in the state shown in Figure 7, manually move the X-Y table 13 so that the chip 9 is placed inside the window frame 12 as shown in the state shown in Figure 7. Move the window frame 12. Window frame 12 (!, y) and chip size (x
, ya) and the street size (xbs yb), as shown in FIG. 7, it is necessary to satisfy the following condition: Xa<X<2xa+zb ya<y<2ya+yb. If it is not satisfied, two chips 9 will be placed within the window frame 12, resulting in two chip center coordinates, and the chip 9 will be left behind, or the row or column will be rotated.

After confirming that the chip 9 has entered the window frame 12, the die
Press the start switch (not shown) on the bonding device to start machine operation. When a start switch (not shown) is pressed, the recognition device 7 determines whether the product is good or bad, performs center position processing as described later, and sends center position data.

If the chip is defective, the X-Y table 13 is moved to position the next chip 9a.

The moving distance is X direction xa+xb=x and y direction y
Let a+yb=Y.

Here, an outline of center position recognition will be given based on FIGS. 8 and 9. As shown in FIG. 8, the image signal sent from the recognition camera 3 is processed by the binarization converter 16, and the input image processing signal is converted into a value (T to become, to do. The binarized data is sent to the image calculation processing section 16 for projection processing, and is classified into X-direction and y-direction projection data as shown in FIG. The classified data is set to a certain value (T
After slicing with the chip size (xa, ya) and the center in the X direction (xc
) and the center in the y direction (yC) to find the center of the chip (c, c).

After receiving the position data after recognition is completed, set the X-Y table 13.
is moved to the suction position (the position where the suction tool 14 is lowered). At the completion signal indicating that the positioning at the suction position is completed, the suction tool 14 is moved upward by 92〉-1 and lowered (■), the chip 9 is suctioned, and the suction tool 14 is raised (■).

■The suction tool 14.

The chip 9 is mounted by performing operations ① and ②.

The X-Y table 13 is positioned at the recognition position to recognize the next step 9a at the timing when the suction tool 14 is raised. Upon receiving the positioning end signal of the XY table 13, the recognition device 7 moves on to recognition of the next chip 9a. After repeating the above positioning, recognition, and head operation, and returning to the first position (the position recognized for the first time after replacing the wafer ring 2), there are no more good chips on the wafer ring 2. 9 Move on to the operation of replacing the ever ring 2. At this time, if there is no replacement wafer ring 2a, the operation ends. When there is a Ueno-ring 2a for replacement, the process moves to the positioning operation of the X-Y table 13 and proceeds to the ninth cycle.

Problems to be Solved by the Invention However, in the above-described operation, the positional relationship between the wafer 1 and the wafer ring 2 is not uniquely determined. When this occurs, it is extremely rare for the state shown in (■) in Figure 7 to occur, and automation is difficult. However, if you place several chips, for example 4 steps, within the window frame, it is possible to place one chip completely within the window frame when you move it to the first position, and use that chip as a reference. Although recognition is possible, there is a problem in that as the number of chips increases, it takes more time to process the recognition for each chip.

Also, when trying to recognize various types of chips with the same recognition device, the chip size (6t 1'6) and street size (xb, yb) are different in size, so it is not possible to recognize them automatically at the first recognition. In order to do this, it is necessary to fit several chips within the window frame, but this not only takes time to recognize, but also makes it difficult to recognize small chips even if the window frame is made smaller. If the magnification of the lens is fixed, even if a large chip and a small chip have the same area ΔS, the small chip will have a large amount of information, and errors in defect determination and center position will easily occur, resulting in poor accuracy.

In view of the above-mentioned conventional drawbacks, the present invention automatically corrects the filtering position when positioned at the first position, guides the chip so that it does not touch the window frame, automates the die bonding equipment, and achieves multiple functions. It is an object of the present invention to provide a recognition device with a position correction function that can shorten recognition time and improve accuracy by recognizing one chip even if the type of chip is different.

Means for Solving the Problems In order to achieve the above object, the present invention includes a recognition device that includes a conversion section that converts an image signal from a capture device into binary levels of white and black, and a conversion section that converts an image signal from a capture device into a binary level of black and white, and a conversion section that converts the converted digital signal an image calculation processing unit that calculates and calculates the center using pattern information input in advance, and a monitor controller that monitors image information of the recognition target and processed information that has been subjected to image calculation processing)
a control panel for teaching image information, an interface section with external equipment, and a data inversion section that selectively inverts the logic of the binarized image signal input to the image processing section. It is characterized by having the following.

Effect By the above formation, when the chip is not in the window frame, by inverting the logic of the binarized image signal that enters the image processing unit, the recognition target is changed from a chip to a street. Based on the obtained data, position correction data can be created to correct the position of the chip within the window frame, and the position of the chip at the first position can be automatically corrected.

EXAMPLE An example of the present invention will be described below with reference to the drawings.

Note that parts that are the same as those of the conventional example are indicated by the same numbers.

FIG. 1 shows an internal configuration diagram of a recognition device with a position correction function according to an embodiment of the present invention. In FIG. 1, reference numeral 21 is an inversion circuit for binary data. Conventionally, the mill 11 was set to "black" 0, but by passing it through this inversion circuit 21, white "0" is set.
Set it to “black” 1.

Reference numeral 22 denotes a data switching signal for switching the data input to the image calculation processing section 16.

Regarding the recognition device with a position correction function configured as above, the difference in its operation from the conventional example explained in FIG. 6 (section A in FIG. 6) will be explained below based on the flowchart in FIG. 2. .

First, replace the wafer ring 2 and replace the X-Y table 13.
is positioned at the first position. Next, a recognition operation is performed to obtain the center data of the chip. If the center data cannot be obtained at this time, it is known that the state is in the state I in FIG. do. The image calculation processing unit 16 performs projection processing, classifies the data into X-direction and y-direction projection data as shown in FIG. 3, slices the classified data at a certain value (Ti(L)), and removes noise. Do it.X from street size (xb, yb)
The street center in the direction (!,) and the street center in the y direction (y8) are found, and the final data, the street center (8C), is found. Next, the position 1d correction amount is determined from the street center (sc) data, chip size (!6t 1a), and street size (xb, yb).

xa ten xb
Determine whether the chip is good or bad and find the center of the chip.

If the center of the chip is aligned, the machine will start operating.

As described above, according to this embodiment, by processing (inversion processing) the information that enters the image calculation processing unit, if the chip center cannot be recognized, the street center is determined, thereby determining the tip size, the street size, and the recognition center. The position of the chip can be corrected from position to position, and automatic position correction can be realized.

Effects of the Invention As described above, according to the present invention, by inverting the logic of the binary level of the image signal, it is possible to find the center between the objects to be recognized and automatically guide the chip within the window frame. The practical effects are significant.

[Brief explanation of the drawing]

FIG. 1 is an internal configuration diagram of a recognition device according to an embodiment of the present invention, and FIG. 2 is a flowchart of a portion of the same operation that differs from a conventional example.
Figure 3 is a projection data diagram of an inverted binary screen within the recognition range, Figure 4 is a schematic configuration diagram of the die bonding device, Figure 6 is an internal configuration diagram of a conventional recognition device, and Figure 6 is an outline of its operation. Flowchart, Figure 7 is a diagram of the positional relationship between the window frame and the chip, Figure 8
The figure shows an image signal and a binary image, and FIG. 9 shows a binary screen within the recognition range and projection data. 3...Recognition camera, 7...Recognition device, 8...Monitor TV 1.9...Terra 7", 10.11. ...Street, 12...
...Window frame, 16...Binarization conversion section, 16.
...Image calculation processing unit, 17...Monitor control section, 18...Operation panel, 19...
. . . Interface section, 20 . . . Central control section, 21 . . . Inverting circuit, 22 . . . Data switching signal. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
Figure 7' - Shun Ranran! Figure 2 Figure 3 Figure 5 Figure 7 Figure 8 Figure 9 Mi

Claims (1)

[Claims] A capture device that captures the state of the recognition target, a recognition device that processes image information from the capture device, and a monitor that monitors the image information of the recognition target sent to the recognition device and processing information of the recognition device. Consisting of equipment,
The recognition device includes a conversion unit that converts the image information sent from the capture device into a binary level of black and white, and an image processing unit that calculates the converted digital information and calculates the center using pattern information input in advance. an arithmetic processing unit, a monitor control unit that monitors the image information of the recognition target and the processed information processed by the image arithmetic processing, an operation panel for teaching the image information, an interface unit with external equipment, and the image arithmetic processing unit. 1. A recognition device with a position correction function, comprising: a data inversion section that selectively inverts the logic of binarized image information input to the section.