JP7245503B2 - BONDING WIRE SHAPE DETECTION DEVICE AND WIRE SHAPE DETECTION METHOD - Google Patents

Description

The present invention relates to a wire shape detection device and method for bonding wires formed by a wire bonding apparatus.

2. Description of the Related Art The height of bonding wires (hereinafter referred to as wires) connecting pads of a semiconductor chip and leads of a substrate is measured. As a method of measuring the wire height, there is a method of taking an image of the wire and measuring the focused height of the edge of the wire in the image (see, for example, Patent Document 1), and a method of measuring the reflected light from the vertex of the wire loop. There is a method of capturing images with a first imaging device directly above the wire and a second imaging device arranged diagonally above the wire, and detecting the height and position of the apex of the wire loop from the two images (see, for example, Patent Document 2). Proposed.

However, in the conventional techniques described in Patent Literatures 1 and 2, the background of the wire becomes bright due to the lighting, and thus the contrast of the wire image is lowered, and the detection accuracy may be lowered.

Therefore, the wire is illuminated with a ring-shaped illuminator, the depth of focus is made shallow, and a dark part appears in the center of the wire image at the focal point. A detection method has been proposed (see Patent Document 3, for example).

JP 2017-45752 A JP 2006-292647 A Japanese Patent No. 3235009

By the way, when a wire is illuminated by a ring-shaped illuminator, in a portion where the wire extends in a substantially horizontal direction, an image is obtained in which the vicinity of the center line of the wire is dark and the edges at both ends of the wire in the width direction are bright at the focal point. Conversely, in a portion where the wire is inclined, an image may be obtained in which the vicinity of the center line of the wire is bright and the edges at both ends in the width direction of the wire are dark. For this reason, in the conventional technique described in Patent Document 3, there are cases where the accuracy of detecting the three-dimensional shape of the wire as a whole is lowered when the wire has an inclined portion.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to accurately detect the three-dimensional shape of a wire.

The wire shape detection apparatus of the present invention comprises an image pickup device for picking up an image of a wire connecting a pad of a semiconductor chip and a lead on a substrate, a ring-shaped illuminator arranged around the image pickup device for illuminating the wire, and an image pickup device. A wire shape detection device comprising: a control unit that causes the device to perform an imaging operation and detects the shape of the wire based on the image of the wire captured by the imaging device, wherein the control unit detects the shape of the wire captured by the imaging device. The image is processed to acquire a plurality of segmented images with different imaging focal heights for each of a plurality of segments of a minute length along the direction in which the wire extends, and a plurality of segmented images with different imaging focal heights in one segment. An image with high contrast is selected as a candidate image for one segment, the selected candidate image is compared with the predicted image, and if the candidate image matches the predicted image, the imaging focal height of the selected candidate image is adjusted. It is characterized by detecting as a wire height in one section and detecting a candidate image as a wire image in one section.

In the wire shape detection apparatus of the present invention, the control unit sequentially detects the wire height and the wire image from one end section of the wire to the other end section. to detect the shape of the wire.

In the wire shape detection device of the present invention, the control unit changes the focal height of the imaging device in stages to capture an entire image of the wire for each different imaging focal height, and converts the captured entire image of the wire into an image of the wire. A segmented image generation operation for generating a plurality of segmented images by dividing into a plurality of segmented images having minute lengths along the extending direction is performed for each captured whole image, and a plurality of segmented images having minute lengths along the direction in which the wire extends is performed. A plurality of segmented images having different imaging focal heights for each segment may be acquired.

In the wire shape detection device of the present invention, the control unit may capture an entire image of the wire for each focal depth range of the imaging device.

The wire shape detection method of the present invention comprises: an imaging device for capturing an image of a wire connecting a pad of a semiconductor chip and a lead on a substrate; a ring-shaped illuminator disposed around the imaging device for illuminating the wire; A wire shape detection method used in a wire shape detection device comprising: a control unit that causes the device to perform an imaging operation and detects the shape of the wire based on the image of the wire captured by the imaging device, wherein the image is captured by the imaging device a segmented image acquisition step of processing the image of the wire to acquire a plurality of segmented images with different imaging focus heights for each of a plurality of segments of a minute length along the direction in which the wire extends; An image with a large contrast is selected as a candidate image for one segment from among a plurality of segmented images with different degrees of contrast, and the selected candidate image is compared with the predicted image. a detection operation step of detecting the imaging focal height of the candidate image as the wire height in one section, and detecting the candidate image as the wire image in the one section.

In the wire shape detection method of the present invention, the wire shape is detected by performing the detection operation step sequentially from one end section of the wire to the other end section, and acquiring the wire height and wire image in each section. It may have a detection step.

In the wire shape detection method of the present invention, the segmented image acquisition step includes a whole image capturing step of capturing an entire image of the wire for each different image capturing focus height by changing the focus height of the image capturing device step by step; A segmented image generating operation for generating a plurality of segmented images by dividing the entire image of the wire into a plurality of segments with minute lengths along the direction in which the wire extends is performed for each captured whole image, and and a segmented image generation step of acquiring a plurality of segmented images with different imaging focal heights for each of the plurality of segments of minute length.

In the wire shape detection method of the present invention, the overall image capturing step may capture an overall image of the wire for each focal depth range of the imaging device.

The present invention can accurately detect the three-dimensional shape of a wire.

1 is a system diagram showing the configuration of a wire shape detection device according to an embodiment; FIG. It is a flow chart which shows operation of a wire shape detection device of an embodiment. It is a flow chart which shows operation of a wire shape detection device of an embodiment. It is explanatory drawing which shows the operation|movement of the whole image imaging step and the division image generation step in the wire shape detection apparatus of embodiment. FIG. 10 is an explanatory diagram showing the operation of the wire shape detection step of division number 1 in the wire shape detection device of the embodiment; It is explanatory drawing which shows the operation|movement of the wire shape detection step of the division number n in the wire shape detection apparatus of embodiment. It is explanatory drawing which shows the operation|movement of the wire shape detection step of the division number m in the wire shape detection apparatus of embodiment. It is an explanatory view showing a three-dimensional shape of a wire detected by the wire shape detection device of the embodiment. FIG. 9 is an explanatory diagram showing another example of a predicted image of the wire shape detection device of the embodiment;

<Configuration of wire shape detection device>
A wire shape detection device 100 according to an embodiment will be described below with reference to the drawings. As shown in FIG. 1, the wire shape detection device 100 includes an imaging device 16, a ring-shaped illuminator 17, and a control section .

The imaging device 16 is composed of a CCD camera or the like. The imaging device 16 is arranged above the semiconductor chip 15 mounted on the substrate 70 and captures an image of the wires 10 connecting the pads 50 of the semiconductor chip 15 and the lead frames 60 of the substrate 70 .

A ring-shaped illuminator 17 is arranged around the imaging device 16 and illuminates the wire 10 obliquely from above. The ring-shaped illuminator 17 is, for example, a large number of LEDs arranged in a ring.

The imaging device 16 and the ring-shaped illuminator 17 are connected to the controller 18 . The control unit 18 is a computer including a CPU 18a for processing information therein and a memory 18b. The control unit 18 adjusts the focal height of the imaging device 16 and causes the imaging device 16 to perform an imaging operation. Further, the control unit 18 detects the shape of the wire 10 based on the image of the wire 10 captured by the imaging device 16 . In addition, in FIG. 1, the dashed-dotted line shows a signal line.

<Operation of Wire Shape Detector>
Next, the operation of the wire shape detection device 100 and the wire shape detection method will be described with reference to FIGS. 2 to 9. FIG.

<Entire image capturing step>
The control unit 18 changes the focal height of the imaging device 16 step by step and executes an entire image imaging step of imaging the entire images 41 to 46 (see FIG. 4) of the wire 10 for each different imaging focal height. Here, the imaging focal height is the focal height or the in-focus height of the imaging device 16, and is stepped for each focal depth range Hf (see FIG. 4A) where the imaging device 16 is in focus. is set In the following description, the imaging focal height is set to six stages of imaging focal height numbers (1) to (6), as indicated by the dashed-dotted line with (1) to (6) in FIG. 4(a). described as being Broken lines above and below imaging focal height numbers (1) to (6) shown in FIG. 4A indicate focal depth ranges Hf of the imaging height numbers (1) to (6).

The control unit 18 sets the imaging focus height number (N) to 1 in step S101 of FIG. Then, the focal height of the imaging device 16 is adjusted to the imaging focal height number (1) shown in FIG. 4(a). Then, the imaging device 16 is focused within the focal depth range Hf between the broken lines above and below the imaging focus height number (1). In this state, the control unit 18 captures the entire image 41 of the wire 10 as shown in step S102 of FIG. As shown in FIG. 4, the entire image 41 of the wire 10 that has been imaged is the image of the second bonding point 12 and the wire 10 with high contrast between the focal depth range Hf of the imaging focal height number (1), and the focal point. and a low-contrast blurred image indicated by a dashed line outside the depth range Hf. Since the wire 10 is inclined in the vicinity of the second bonding point 12, when the wire 10 is illuminated by the ring-shaped illuminator 17 shown in FIG. is a single-line image that is bright near the center line and dark on both sides in the width direction. Then, in step S103 of FIG. 2, the control unit 18 stores the captured whole image 41 in the memory 18b. In practice, the image gradually changes from a focused image to a blurred image as the distance from the depth of focus range Hf increases. are divided into two types of images.

The control unit 18 increments the imaging focus height number (N) by 1 in step S104 in FIG. 2 to 2, and returns to step S102 in FIG. At the same time, the overall image 42 of the wire 10 is picked up, and the overall image 42 is stored in the memory 18b in step S103 of FIG. As shown in FIG. 4, the whole image 42 of the wire 10 captured is outside the high-contrast image of the wire 10 between the depth of focus range Hf of the imaging focus height number (2) and the depth of focus range Hf. It consists of a low-contrast, blurred image indicated by the dashed line. Since the wire 10 in this range is also inclined, the entire image 42 of the wire 10 captured is also a single-line image as shown in FIG.

Similarly, every time the imaging focus height number (N) is incremented by 1 in step S104 in FIG. The whole images 43 to 46 of the wire 10 are sequentially picked up and stored in the memory 18b.

The whole image 43 of the wire 10 with the imaging focus height number (3) is one of the wires 10 with high contrast within the depth of focus range Hf, similar to the whole image 42 of the wire 10 with the imaging focus height number (2). It is composed of a main line image and a low-contrast blurred single line image indicated by a broken line outside the depth of focus range Hf.

In the imaging focus height number (4), the first bonding point 11 where the pressure-bonded ball is formed falls within the depth-of-focus range Hf. An outline image 11a of the crimp ball at the first bond point 11 is included in addition to the 10 single line images.

In the focal depth range Hf of the imaging focal height number (6), there is a portion where the wire 10 extends substantially horizontally. When this portion of the wire 10 is illuminated by the ring-shaped illuminator 17 shown in FIG. 1 and an image of the wire 10 is captured by the imaging device 16, two lines with high contrast are bright at both ends in the width direction of the wire 10 and dark near the center line. becomes an image of In addition, since the wire 10 is inclined downward on both sides in the length direction of the wire 10 in this portion, a single line image with high contrast is obtained. In addition, the overall image 46 includes a low-contrast blurred single-line image indicated by the dashed line outside the depth-of-focus range Hf.

The entire image 45 of the imaging focus height number (5) is a single line image of the wire 10 with high contrast between the depth of focus range Hf and a single line with low contrast indicated by a broken line outside the depth of focus range Hf. It includes a blurred image and a low-contrast image (indicated by a dashed line) having the same shape as the double-line image and the single-line image included in the imaging focus height number (6).

In this way, every time the imaging focus height number (N) is incremented by 1 in step S104 in FIG. Then, the whole images 41 to 46 of the wire 10 are sequentially picked up and stored in the memory 18b. Then, in step S105, it is determined whether or not the imaging focus height number (N) exceeds the final number Nend. In this embodiment, since the imaging focus height number N is set to six levels, Nend is 6, and six whole images 41 to 46 are picked up in the whole image pickup step.

Then, when N becomes 7 and it is determined as YES in step S105 of FIG. Then, the entire image capturing step is terminated, and the process proceeds to step S106 in FIG.

<Segmented image generation step>
The control unit 18 divides the captured whole images 41 to 46 of the wire 10 into a plurality of segments with minute lengths along the direction in which the wire 10 extends to generate a plurality of segmented images. A segmented image generation step is performed for each of the images 41 to 46 to obtain a plurality of segmented images with different imaging focal heights for each of a plurality of segments of minute lengths along the direction in which the wire 10 extends. As shown in FIG. 4( a ), the division is made by assigning division number (1) to the first bonding point 11 side, which is one end, and division number (Mend) to the second bonding point side, which is the other end. divided into pieces. In the following description, a segmented image with an imaging focus height number (N) and a segmentation number (M) will be referred to as a segmented image (N, M).

The control unit 18 sets the imaging focus height number (N) and the division number (M) to the initial values of 1 in step S106 of FIG. The control unit 18 cuts out the segmented image (1, 1) of the segmentation number (1) of the whole image 41 of the imaging focal height number (1) in step S107 in FIG. 2, proceeds to step S108 in FIG. store in

The control unit 18 increments the division number (M) by 1 in step S109 of FIG. 2 and repeats steps S107 to S109 until the division number (M) reaches the final number Mend. As a result, Mend segmented images of each segmented image (1, 1) to segmented image (1, Mend) in segment number (1) to segment number (Mend) of the entire image 41 of imaging focus height number (1) cut out.

When the division number (M) exceeds the final number Mend, the control unit 18 determines YES in step S110 in FIG. 2, proceeds to step S111 in FIG. is reset to 1, and in step S112, the imaging focal point height number (N) is incremented by 1 to be 2, and the process proceeds to step S113. Since the imaging focus height number (N) is 2 and does not exceed the final number Nend of 6, the control unit 18 determines NO in step 113 and returns to step S107. Then, steps S107 to S109 are repeated until the section number (M) reaches the final number Mend, and each section image ( 2, 1) to segmented image (2, Mend) are cut out as Mend segmented images.

In this way, the control unit 18 cuts out segmented images (N, M) for each imaging focus height number N, and stores each segmented image (N, M) in the memory 18b. Then, in step S113, it is determined whether or not the imaging focus height number (N) exceeds the final number Nend. In this embodiment, since the imaging focal height number (N) is set to six levels, Nend is 6. , segmented image (2, 1) to segmented image (2, Mend), . . . segmented image (6, 1) to segmented image (6, Mend).

Then, when N becomes 7 and it is determined YES in step S113 of FIG. 2, the control unit 18 determines that all segmented images have been obtained, terminates the segmented image generation step, and proceeds to step S113 of FIG. Proceed to S114.

Note that the whole image capturing step and the segmented image generation step constitute the segmented image acquisition step.

<Wire shape detection step>
After setting the partition number M to the initial value of 1 in step S114 in FIG. 2, the control unit 18 proceeds to step S115 in FIG. 3 to set the predicted image (M) corresponding to the partition number (M). Then, in step S116 of FIG. 3, Nend segmented images (1,M) to (Nend,M) of segmented images (1,M) to (Nend,M) having imaging focal height numbers (N) of segmentation number (M) are (1) to (Nend). Read out from the memory 18b, and in step S117 of FIG. 3, the image with the maximum contrast among the Nend segmented images (1, M) to (Nend, M) is selected as the candidate image (M) for the segment number (M). select.

Then, the candidate image (M) selected in step S118 in FIG. 3 is compared with the predicted image (M), and if they match, the process proceeds to step S120 in FIG. ) is detected as the wire image of the division number (M) and stored in the memory 18b. Also, in step S121, the height of the imaging focus height number (N) of the candidate image (M) is detected as the wire height and stored in the memory 18b (detection operation step).

If the control unit 18 determines NO in step S118 of FIG. 3, the control unit 18 proceeds to step S119 of FIG. Execute.

In this way, the control unit 18 performs the above detection operation steps from the section number (1) on the first bond point 11 side, which is the one end side, to the section number (Mend) on the second bond point side, which is the other end side. This is repeated for each section number (M), and the wire image and wire height in each section from section number (1) to section number (Mend) are acquired and stored in the memory 18b. Thereby, the three-dimensional shape of the wire 10 is detected. This is the wire shape detection step.

<Acquisition of wire images and wire heights in division numbers (1), (n), and (m)>
The detection of wire images and wire heights in division numbers (1), (n), and (m) will be described in detail below as an example with reference to FIGS. 5 to 7. FIG.

<Example of division number (1)>
As shown in FIG. 5, in the area of division number (1), the wire 10 rises vertically from the first bonding point 11 and bends in the horizontal direction. Therefore, since the wire 10 is inclined, when the wire 10 is illuminated by the ring-shaped illuminator 17 shown in FIG. Therefore, in step S115 of FIG. 3, the control unit 18 sets the single-line image 31 as shown in FIG. 5 as the prediction image (1) of the division number (1), and proceeds to step S116 of FIG.

The control unit 18 reads the six segmented images (1, 1) to (6, 1) shown in FIG. 5 from the memory 18b in step S116 of FIG. As shown in FIG. 5, in the section number (1) of M=1, the section image (6, 1) of the imaging focus height number (6) has a single line image 21 with high contrast. A segmented image (4, 1) with a number (4) has an outline image 11a of a pressure-bonded ball of the first bond point 11. FIG. The contour image 11a of the compression ball is an image with low contrast. Clear images do not exist in the segmented images (1,1) to (3,1), (5,1) of the other imaging focal height numbers (1) to (3), (5).

Then, in step S117 of FIG. 3, the control unit 18 selects the segmented image (6,1) having the maximum contrast among the six segmented images (1,1) to (6,1) as shown in FIG. ) as the candidate image (1). Then, the process proceeds to step S118 in FIG. 3 to compare the candidate image (1) and the predicted image (1). Since both the candidate image (1) and the predicted image (1) are single-line images, the control unit 18 determines YES in step S118 of FIG. 3, proceeds to step S120 of FIG. is detected as a wire image of division number (1) and stored in the memory 18b. Also, in step S121 in FIG. 3, the height of the imaging focus height number (6) of the candidate image (1) is detected as the wire height of the division number (1) and stored in the memory 18b.

Then, the control unit 18 increments the division number (M) by 1 to make M=2 in step S122 of FIG. In this case, the control unit 18 determines NO in step S123 in FIG. 3, returns to step S115 in FIG. 3, executes steps S115 to S121 in FIG. to detect

<Example of division number (n)>
Next, wire 10 extends in a substantially horizontal direction, and detection of the wire shape and wire height of section number (n) in which the image of wire 10 captured by imaging device 16 becomes two lines will be described. Operations similar to those of division number (1) will be briefly described.

In the division number (n), the image of the wire 10 is a two-line image, so the control unit 18 sets the two-line image 32 shown in FIG. 6 as the predicted image (n) in step S115 of FIG.

Then, in step S116, the six divided images (1, n) to (6, n) are read out from the memory 18b. The segmented image (6,n) contains a high-contrast double-line image 22b, the segmented image (5,n) contains a low-contrast double-line image 22a, and the other segmented image (1,n ) to (4, n) do not have clear images. Therefore, the control unit 18 selects the partitioned image (6, n) with the maximum contrast as the candidate image (n) in step S117 of FIG. In this case, since the candidate image (n) is an image with two lines similar to the predicted image (n), the control unit 18 determines YES in step S118 of FIG. (n) is detected as a wire shape and stored in the memory 18b, and in step S121 of FIG. , and stored in the memory 18b.

If the control unit 18 determines NO in step S118 in FIG. 3, the process proceeds to step S119 in FIG. Select the image as a candidate image (n). In this example, the only image with a height number adjacent to the segmented image (6, n) is the segmented image (5, n). Select and return to step S118 in FIG. Since the segmented image (5, n) is a two-line image similar to the predicted image (n), the control unit 18 determines YES in step S118, and in steps S120 and S121 of FIG. is detected and stored in the memory 18b.

<Example of division number (m)>
Next, detection of the wire shape and wire height of the division number (m) in which the wire 10 is inclined and the image of the wire 10 captured by the imaging device 16 is a single line will be described. Operations similar to those of division number (n) will be briefly described.

In the division number (m), the image of the wire 10 is a single-line image, so the controller 18 sets the single-line image 33 shown in FIG. 7 as the predicted image (m) in step S115 of FIG.

Then, in step S116, six images (1, m) to (6, m) are read from the memory 18b. Segmented images (4,m)-(6,m) each include single-line images 23a-23c. The image 23b of the segmented image (5,m) has the highest contrast, the image 23c of the segmented image (6,m) has a lower contrast than the image 23b of the segmented image (5,m), and the image 23b of the segmented image (4,m) has the highest contrast. The image 23a in m) has the lowest contrast. There is no clear image in the other segmented images (1,m)-(3,m). Therefore, the control unit 18 selects the segmented image (5,m) with the maximum contrast as the candidate image (m) in step S117 of FIG. In this case, since the candidate image (m) is a single-line image similar to the predicted image (m), the control unit 18 determines YES in step S118 of FIG. (m) is detected as a wire shape and stored in the memory 18b, and in step S121 of FIG. , and stored in the memory 18b.

As in the case of the section number (n), if the control unit 18 determines NO in step S118 in FIG. Select images with adjacent height numbers as candidate images (m). In this example, there are two images with height numbers adjacent to segmented image (5, m): segmented image (4, m) and segmented image (6, m). The control unit 18 selects the segmented image (6, m), which has the greater contrast between the segmented image (4, m) and the segment number (6, m), as the candidate image (m), and proceeds to step S118 in FIG. return. Since the segmented image (6,m) is a single-line image similar to the predicted image (m), the control unit 18 determines YES in step S118, and in steps S120 and S121 of FIG. is detected and stored in the memory 18b.

As described above, the control unit 18 detects the wire shape and the wire height for each of the division numbers (1) to (Mend) and stores them in the memory 18b. When M exceeds Mend, YES is determined in step S123 of FIG. 3, and the wire shape detection step ends.

As shown in FIG. 8(a), when the wire shapes of section number (1) to section number (Mend) acquired by the control unit 18 are arranged in order of section number (M), an entire detection image 40 of the wire 10 is obtained. is obtained. Thereby, the planar shape of the wire 10 can be detected. Also, if the wire heights of each section number (M) are arranged in the order of the section number (M), the vertical shape of the wire 10 can be detected as shown in FIG. 8(b). Then, the three-dimensional shape of the wire 10 can be detected based on FIGS. 8(a) and 8(b).

As described above, in the wire shape detection device 100 of the embodiment, when the wire 10 is imaged by the imaging device 16 using the ring-shaped illuminator 17 shown in FIG. The three-dimensional shape of the wire 10 can be detected with high accuracy even when the image includes an inclined portion that is a single line image.

In the description of the embodiment, the imaging focus height number (N) is divided into 6 stages, but the imaging device 16 with a narrow focal depth range Hf may be used, or the depth of focus may be reduced by darkening the illumination. A more detailed three-dimensional shape may be detected by dividing into a larger number of stages by reducing the range Hf.

In addition, in the embodiment, two types of images, a double-line image and a single-line image, are used as predicted images. Between the single-line image, the inclined double-line image 34 or the chevron-line image 35 may be used as the prediction image.

Also, in the wire shape detection step, the wire shape and wire height detected in the section number (k) may be referenced to detect the wire shape and wire height in the next section number (k+1). In this way, by referring to the wire shape and wire height of the division number (k) for which the wire shape and wire height were detected earlier, it is possible to suppress erroneous detection in the next division number (k+1). can.

10 wire, 11 first bond point, 11a contour image, 12 second bond point, 15 semiconductor chip, 16 imaging device, 17 ring-shaped illuminator, 18 control unit, 18a CPU, 18b memory, 21, 22a, 22b, 23a 23c, 31-35 images, 40 overall detection images, 41-46 overall images, 50 pads, 60 lead frames, 70 substrates, 100 wire shape detectors.

Claims (8)

an imaging device that captures an image of a wire that connects the pad of the semiconductor chip and the lead of the substrate;
a ring-shaped illuminator arranged around the imaging device to illuminate the wire;
a control unit that causes the imaging device to perform an imaging operation and detects the shape of the wire based on the image of the wire captured by the imaging device;
A wire shape detection device comprising:
The control unit
processing the image of the wire captured by the imaging device to obtain a plurality of segmented images with different imaging focal heights for each of a plurality of segments of a minute length along the direction in which the wire extends;
selecting an image having a large contrast as a candidate image for one of the segments from among a plurality of segmented images having different imaging focal heights in the one segment; comparing the selected candidate image with a predicted image; is matched with the predicted image, detecting the imaging focal height of the selected candidate image as the wire height in one of the sections, and detecting the candidate image as a wire image in one of the sections;
A wire shape detection device characterized by: The wire shape detection device according to claim 1,
The control unit sequentially detects the wire height and the wire image from the section at one end of the wire toward the section at the other end, and detects the wire height and the wire image in each section. to detect the shape of the wire;
A wire shape detection device characterized by: The wire shape detection device according to claim 2,
The control unit
Imaging an entire image of the wire for each different imaging focal height by changing the focal height of the imaging device step by step,
A segmented image generating operation for generating a plurality of segmented images by dividing the captured whole image of the wire into a plurality of segments of minute lengths along the direction in which the wire extends is performed for each of the captured whole images. obtaining a plurality of segmented images with different imaging focal heights for each of the plurality of segments of minute length along the direction in which the wire extends;
A wire shape detection device characterized by: The wire shape detection device according to claim 3,
The control unit
Capturing the entire image of the wire for each focal depth range of the imaging device;
A wire shape detection device characterized by: An imaging device that captures an image of a wire that connects a pad of a semiconductor chip and a lead of a substrate, a ring-shaped illuminator that is arranged around the imaging device and illuminates the wire, and the imaging device performs an imaging operation. A wire shape detection method used in a wire shape detection device comprising:
A segmented image acquisition step of processing the image of the wire captured by the imaging device to acquire a plurality of segmented images with different imaging focal heights for each of a plurality of segments of a minute length along the direction in which the wire extends;
selecting an image having a large contrast as a candidate image for one of said segments from among a plurality of segmented images with different imaging focal heights in said one segment, comparing said selected candidate image with a predicted image, and said candidate image; a detection operation step of detecting the imaging focal height of the selected candidate image as a wire height in one of the sections and detecting the candidate image as a wire image in one of the sections when the is matched with the predicted image. and,
A wire shape detection method characterized by comprising: The wire shape detection method according to claim 5,
Wire shape detection for detecting the shape of the wire by sequentially performing the detection operation step from the section at one end of the wire toward the section at the other end, and acquiring the wire height and the wire image in each section. a step;
A wire shape detection method characterized by comprising: The wire shape detection method according to claim 6,
The segmented image acquisition step includes:
an overall image capturing step of capturing an entire image of the wire for each different imaging focus height by changing the focal height of the imaging device stepwise;
A segmented image generating operation for generating a plurality of segmented images by dividing the captured whole image of the wire into a plurality of segments of minute lengths along the direction in which the wire extends is performed for each of the captured whole images. a segmented image generation step of acquiring a plurality of segmented images with different imaging focal heights for each of the plurality of segments having minute lengths along the direction in which the wire extends;
A wire shape detection method, comprising: The wire shape detection method according to claim 7,
The whole image capturing step includes:
Capturing the entire image of the wire for each focal depth range of the imaging device;
A wire shape detection method characterized by:

Images