JP6132682B2 - Key pattern detection method - Google Patents

Description

  The present invention relates to a key pattern detection method for detecting a key pattern on a workpiece.

  A workpiece such as a wafer having a device formed on the surface thereof is processed by, for example, a cutting apparatus including an annular cutting blade or a laser processing apparatus including a processing head for laser irradiation.

  Prior to processing the workpiece, an alignment process is performed to identify and adjust the positional relationship between the processing means such as a cutting blade or a processing head and the workpiece (see, for example, Patent Document 1).

  In the alignment process, first, a characteristic pattern of a registration workpiece is captured by a camera and set as a key pattern. Next, a processing workpiece to be processed is imaged, and a pattern that matches the key pattern is detected from the captured image.

  Thereafter, the positional relationship between the processing means and the workpiece is specified based on the detected pattern position, and the positional relationship suitable for processing is adjusted. Thus, by adjusting the positional relationship between the processing means and the workpiece, the workpiece can be appropriately processed along the scheduled processing line.

JP-A-7-106405

  By the way, the workpiece has scratches and dirt, and the pattern itself has a slight error. Therefore, it is not possible to expect complete matching between the set key pattern and the pattern to be detected.

  For this reason, in the above-described alignment processing, pattern matching that detects a pattern having the strongest correlation (highest correlation) with the key pattern is adopted from the patterns included in the captured image obtained by capturing the processing workpiece.

  In order to increase the accuracy of this pattern matching, it is necessary to sufficiently remove the influence of scratches and dirt on the surface and strengthen the correlation between the set key pattern and the pattern to be detected. Therefore, when imaging a workpiece, the amount, color, direction, etc. of the irradiated light are adjusted so that scratches and dirt on the surface do not appear in the captured image.

  For example, if the amount of light applied to the surface of the workpiece is increased from an appropriate amount, scratches on the surface can be overexposed so that they are not reflected in the captured image. In this case, the picked-up image picked up by increasing the amount of light to be irradiated is registered as a key pattern, and the processing workpiece is picked up with the same amount of light.

  However, in the captured image captured by such a method, the pattern itself including the key pattern may be overexposed, and there is a problem that high-precision pattern matching cannot always be realized.

  Further, in pattern matching of a workpiece whose back side is ground, the workpiece may be imaged from the back side using an infrared camera. In this case, the grinding mark on the back surface side appears in the captured image together with the pattern on the front surface side.

  Since the direction of the grinding mark varies from region to region even in the same workpiece, it is difficult to realize high-precision pattern matching when a captured image in which the grinding mark is reflected is registered as a key pattern.

  The present invention has been made in view of such problems, and an object of the present invention is to provide a key pattern detection method capable of detecting a key pattern with high accuracy.

  According to the present invention, there is provided a key pattern detection method for detecting a key pattern on a workpiece, and a registration step for registering a key pattern image having a first number of pixels formed by imaging a feature point on the workpiece. A workpiece captured image forming step of imaging a workpiece and forming a workpiece captured image having a second field of view with a wider field of view than the key pattern image, and the intensity of each pixel of the key pattern image, A binarization processing threshold value is calculated by calculating a gray level average value obtained by dividing a value obtained by summing the gray levels of each pixel of the workpiece captured image by the sum of the first pixel number and the second pixel number; And a binarization processing step for performing binarization processing on the key pattern image with the threshold for binarization processing set in the threshold setting step. The law is provided.

  According to the present invention, there is also provided a key pattern detection method for detecting a key pattern on a workpiece, wherein a key pattern image having a first pixel number formed by imaging a feature point on the workpiece is registered. A registration step, a workpiece captured image forming step of imaging the workpiece and forming a workpiece captured image having a second field of view with a wider field of view than the key pattern image, and the intensity of each pixel of the key pattern image A first average value calculating step for dividing the sum of the two by the first number of pixels to calculate a key pattern image intensity average value, and calculating the total intensity of each pixel of the workpiece captured image as the second pixel number And the second average value calculating step for calculating the average value of the workpiece captured image intensity, the key pattern image intensity average value calculated in the first average value calculating step, and the second A threshold value setting step for calculating an average value of the workpiece captured image grayscale average value calculated in the average value calculating step and setting it as a threshold value for binarization processing, and the key pattern image set in the threshold value setting step There is provided a key pattern detection method comprising: a binarization processing step for performing binarization processing with a threshold for binarization processing.

  In the key pattern detection method of the present invention, after performing the binarization processing step, the binarization processing is performed on the workpiece captured image with the threshold for binarization processing, and the workpiece captured image is displayed on the workpiece captured image. It is preferable to further include a pattern matching step of calculating a correlation by pattern matching the binarized key pattern image and detecting a region having the strongest correlation as a key pattern.

  According to the present invention, a threshold value suitable for binarization processing is set based on the average value of the intensity of each pixel in the image, so that the scratches and dirt on the workpiece reflected in the image are binarized. It can be removed properly by processing. Thereby, the key pattern can be detected with high accuracy.

It is a perspective view which shows typically the structural example of the laser processing apparatus in which the key pattern detection method is used. It is a figure which shows typically the example of the surface side of a workpiece, and the image imaged. It is a figure which shows an alignment process typically. It is a figure which shows typically the example of the pattern matching implemented in an alignment process. It is a figure which shows a registration step typically. It is a figure which shows a workpiece picked-up image formation step typically. It is a figure which shows a pattern matching step typically.

  FIG. 1 is a perspective view schematically showing a configuration example of a laser processing apparatus in which a key pattern detection method is used. The key pattern detection method of the present invention is not limited to being used in a laser processing apparatus, and may be used in apparatuses such as a cutting apparatus, a water jet processing apparatus, a cleaning apparatus, and a measuring apparatus.

  As shown in FIG. 1, the laser processing apparatus 2 includes a base 4 that supports each component. A housing 6 that covers the base 4 is provided above the base 4. A space is formed inside the housing 6, and a laser irradiation unit 12 including a processing head 8 that irradiates a laser beam and a camera 10 that images the workpiece 11 is accommodated.

  Below the laser irradiation unit 12, a chuck table 14 for sucking and holding the workpiece 11 is provided. The chuck table 14 moves in the X-axis direction (machining feed direction) and the Y-axis direction (index feed direction) horizontal to the upper surface of the base 4 and rotates around the Z-axis.

  A cassette elevator 16 is provided at a corner of the base 4. On this cassette elevator 16, a cassette 18 that houses the workpiece 11 is placed. The cassette elevator 16 is configured to be movable up and down, and adjusts the position of the cassette 18 in the Z-axis direction (height direction) so that the workpiece 11 can be carried out and carried in.

  A touch panel type monitor 20 for setting various conditions is provided on the front surface 6 a of the housing 6. The monitor 20 is connected to a control device 22 (see FIG. 3) that controls each component of the laser processing device 2. The control device 22 controls operations of the machining head 8, the camera 10, the chuck table 14, and the like based on various conditions set through the monitor 20.

  Next, the alignment process performed in the said laser processing apparatus 2 is demonstrated. FIG. 2 is a diagram schematically illustrating an example of a surface side of the workpiece 11 and an image to be captured.

  As shown in FIG. 2, the workpiece 11 is a disk-shaped semiconductor wafer, and includes a central device region 13 and an outer peripheral surplus region 15 surrounding the device region 13. The device region 13 on the surface side of the workpiece 11 is divided into a plurality of regions by streets (scheduled processing lines) 17a and 17b arranged in a lattice pattern, and each region includes a device 19 (19a and 19b) such as an IC. ) Is formed.

  By capturing a part of the device region 13 of the workpiece 11 with the camera 10 of the laser processing apparatus 2, for example, captured images 30a and 30b as shown in FIG. 2 are formed. The captured images 30a and 30b include a pattern P formed by elements and wirings in the devices 19a and 19b.

  In the alignment process, the characteristic pattern P in the devices 19a and 19b is set as the key pattern Pk, and the position of the workpiece 11 is specified. Specifically, using the key pattern Pk, a paralleling step for aligning the street 17a (or street 17b) in parallel with the X-axis direction (or Y-axis direction) and a processing position detection step for detecting a processing position are performed. carry out.

  In the parallelizing step, first, the captured images 30a and 30b of the devices 19a and 19b located at positions separated in the direction parallel to the street 17a are acquired, and the key pattern Pk in each of the captured images 30a and 30b is detected.

  Then, the chuck table 14 is rotated so that the coordinates in the Y-axis direction of the key pattern Pk in the captured images 30a and 30b coincide. Thereby, a straight line connecting the key patterns Pk in the captured images 30a and 30b becomes parallel to the X-axis direction. That is, the street 17a is parallel to the X-axis direction.

  In the machining position detection step, the coordinates of the machining position based on the detected key pattern Pk are calculated based on the key pattern Pk stored in the laser machining apparatus 2 and the coordinate information of the machining position in advance. In addition, the coordinate information mentioned above is acquired based on design information etc., for example.

  As a result of such an alignment process, the workpiece 11 can be appropriately processed by irradiating the laser beam along the streets 17a and 17b in a subsequent processing step. The paralleling step and the processing position detection step are also performed in the Y-axis direction as necessary. In FIG. 2, some of the captured images 30a and 30b are set as the key pattern Pk. However, all of the captured images 30a and 30b may be set as the key pattern Pk.

  FIG. 3 is a diagram schematically showing the alignment process, and FIG. 4 is a diagram schematically showing pattern matching performed in the alignment process. As shown in FIG. 3, before starting the alignment process, the registration work piece 11a for registering the key pattern Pk is imaged, and the key pattern Pk is selected from the formed captured image 30c. The selected key pattern Pk is registered as a key pattern image 32 in the memory 22a of the control device 22.

  When the alignment process is started, the control device 22 causes the camera 10 to capture the processing workpiece 11b to be processed, and stores the obtained captured images 30a and 30b in the memory 22a.

  Further, the control device 22 performs pattern matching for detecting the pattern P having the strongest correlation (highest correlation) with the key pattern image 32 from the patterns P included in the captured images 30a and 30b based on the processing program 22b. To do.

  In this pattern matching, first, as shown in FIG. 4, regions in the captured images 30 a and 30 b are sequentially shifted in the vertical and horizontal directions by a matching frame 34 having the same size (number of pixels) as the key pattern image 32. Are cut out to form a plurality of matching images 36.

  Next, the strength of correlation (high correlation) between each of the plurality of images to be matched 36 formed and the key pattern image 32 is calculated, and the matching image having the strongest correlation with the key pattern image 32 is calculated. 36 is detected as a key pattern Pk. According to this method, the matched image 36 most similar to the key pattern image 32 can be detected as the key pattern Pk.

  By the way, the workpieces 11a and 11b usually have slight scratches and dirt. Therefore, if scratches or dirt appear in the captured images 30a, 30b, and 30c, the key pattern Pk may not be detected properly. Similarly, when imaging the workpiece 11 whose back side is ground by the camera 10 sensitive to infrared rays, grinding marks on the back side are reflected in the captured image 30 and the accuracy of pattern matching is lowered.

  To solve this problem, the present inventor sets a binarization processing threshold for binarizing an image based on the average value of the intensity of each pixel, and the workpiece 11 captured in the image is displayed. It was thought that the key pattern could be detected with high accuracy by appropriately removing the scratches, dirt, etc. in the binarization process. And based on this knowledge, this invention was completed. Embodiments of the present invention will be described below.

(Embodiment 1)
In the present embodiment, a first aspect of the key pattern detection method will be described. The key pattern detection method according to the first aspect includes a registration step, a workpiece captured image formation step, a threshold setting step, a binarization processing step, and a pattern matching step.

  In the registration step, a part or all of the captured image obtained by capturing the characteristic pattern of the workpiece is registered as a key pattern image. In the workpiece captured image forming step, a workpiece captured image is formed by capturing the workpiece with a wider field of view than the key pattern image.

  In the threshold setting step, the gray level of each pixel of the key pattern image and the gray level of each pixel of the workpiece captured image are summed, and the total number of pixels of the key pattern image (first pixel number) and the workpiece captured image By dividing by the sum of the total number of pixels (second pixel number), the average value of the lightness (lightness average value) is calculated. Then, the average value of the calculated shade is set as a threshold value for binarization processing (binarization processing threshold value).

  In the binarization processing step, the key pattern image is binarized based on the threshold for binarization processing. In the pattern matching step, first, the workpiece captured image is binarized based on the threshold for binarization. Next, the correlation between the binarized key pattern image and the pattern included in the binarized workpiece captured image is calculated, and the pattern having the strongest correlation with the key pattern image is detected as the key pattern. To do. Hereinafter, the key pattern detection method according to the present embodiment will be described in detail.

  In the key pattern detection method according to the present embodiment, first, a registration step of registering a part or all of a captured image obtained by capturing a characteristic pattern of a workpiece as a key pattern image is performed. FIG. 5 is a diagram schematically showing the registration step.

  In the registration step, first, the registration workpiece 11 a is sucked and held on the chuck table 14 of the laser processing apparatus 2. And as shown in FIG. 5, the workpiece 11a is imaged with the camera 10 of the laser processing apparatus 2, and the captured image 40a containing the key pattern Pk is obtained.

  In addition, the imaging by the camera 10 may be performed from the front surface side of the workpiece 11a, or may be performed from the back surface side of the workpiece 11a. However, since the key pattern Pk is formed on the front side of the workpiece 11a, it is necessary to use the camera 10 that is sensitive to infrared rays when imaging from the back side of the workpiece 11a.

  Next, the characteristic key pattern Pk included in the captured image 40 a is registered in the memory 22 a of the control device 22 as the key pattern image 42. As shown in FIG. 5, although a part of the captured image 40a is registered as the key pattern image 42 here, the entire captured image 40a may be registered as the key pattern image 42.

  This registration step may be performed by a manual operation of an operator, or may be performed by automatic processing by the control device 22. Note that the registration step may be called a teaching process (teaching process) or the like.

  Further, image processing such as filter processing may be applied to the key pattern image 42 registered in the registration step. For example, edge detection filter processing for detecting an edge (contour) can be applied.

  After the registration step, a workpiece captured image forming step is performed in which the workpiece is captured with a wider field of view than the key pattern image to form a workpiece captured image. FIG. 6 is a diagram schematically illustrating the workpiece captured image forming step.

  In the workpiece captured image forming step, first, the workpiece 11b for processing is sucked and held on the chuck table 14 of the laser processing apparatus 2. Then, as shown in FIG. 6, the workpiece 11 b is imaged by the camera 10 of the laser processing apparatus 2 to obtain a workpiece captured image 40 b having a wider field of view than the key pattern image 42.

  Imaging by the camera 10 may be performed from the front side of the workpiece 11b or from the back side of the workpiece 11b, but needs to be performed in accordance with the registration step. The obtained workpiece captured image 40 b is stored in the memory 22 a of the control device 22.

  After the workpiece captured image forming step, a threshold setting step for setting a threshold (binarization processing threshold) used for the subsequent binarization processing based on the shades of the key pattern image 42 and the workpiece captured image 40b. To implement.

  The key pattern image 42 and the workpiece captured image 40b are an aggregate of a plurality of pixels, and each pixel expresses a color tone (brightness) between “black” and “white” with a plurality of gradations. The shade to be set is set.

The intensity of each pixel is set according to the amount of light received by the camera 10. In general, in an N-bit grayscale image, the intensity of each pixel is expressed in 2 ^N levels (2 ^N gradations). For example, the intensity of “black” is “0” and the intensity of “white” is “2” ^N- 1 ".

  In the threshold setting step of the present embodiment, a threshold used for the later binarization process is set based on this density. Specifically, first, the control device 22 reads the key pattern image 42 and the workpiece captured image 40b stored in the memory 22a.

Then, based on the processing program 22b, an average value of intensity (average intensity value) mean is calculated. The following formula (1) is used to calculate the average value mean. In Expression (1), n ₁ represents the total number of pixels (first number of pixels) of the key pattern image 42, and n ₂ represents the total number of pixels (second number of pixels) of the workpiece captured image 40b. K _i represents the intensity of the i-th pixel included in the key pattern image 42, and W _j represents the intensity of the j-th pixel included in the workpiece captured image 40b.

  Thereafter, the average value mean calculated by the above equation (1) is set as a threshold value used in the subsequent binarization process and stored in the memory 22a. Thus, in the present embodiment, the total value of the intensity of each pixel of the key pattern image 42 and the intensity of each pixel of the workpiece captured image 40b is calculated as the total number of pixels of the key pattern image 42 (first A value obtained by dividing the sum of the number of pixels) and the total number of pixels of the workpiece captured image 40b (second pixel number) is set as a threshold for binarization processing.

  After the threshold setting step, a binarization processing step for binarizing the key pattern image based on the threshold for binarization processing set in the threshold setting step is performed. Here, the binarization processing refers to processing for converting an image expressed in grayscale or color into a monochrome image expressed in two colors of white and black.

  In the binarization processing step, first, the control device 22 reads the key pattern image 42 and the threshold for binarization processing stored in the memory 22a. Then, based on the processing program 22b, the density of each pixel in the key pattern image 42 is compared with a threshold for binarization processing, and each pixel is converted to either “black” or “white”.

  In the present embodiment, since the average value mean of the density is set as a threshold for binarization processing, for example, in the key pattern image 42, pixels with a density less than the average value mean are set to “black”. The pixels having the intensity greater than or equal to the average value mean are converted to “white”. The converted key pattern image 42 is stored in the memory 22a.

  After the binarization processing step, a pattern matching step for detecting a pattern in the workpiece captured image having the strongest correlation with the key pattern image as a key pattern is performed. FIG. 7 is a diagram schematically showing the pattern matching step.

  In the pattern matching step, first, the workpiece captured image 40b is binarized based on the threshold for binarization. Specifically, first, the control device 22 reads the workpiece captured image 40b and the binarization threshold value stored in the memory 22a.

  Then, based on the processing program 22b, the density of each pixel in the key pattern image 42 is compared with a threshold for binarization processing, and each pixel is converted to either “black” or “white”. The converted workpiece captured image 40b is stored in the memory 22a.

  Next, as shown in FIG. 7, the correlation between the binarized key pattern image 42 and the pattern included in the binarized workpiece captured image 40b is calculated, and the key pattern image 42 A pattern having the strongest correlation is detected as a key pattern.

  Specifically, first, the control device 22 reads the workpiece captured image 40b after the binarization processing from the memory 22a, and forms a plurality of matching images (patterns) 44 based on the processing program 22b.

  The plurality of images to be matched 44 are formed by cutting out a region in the workpiece captured image 40b while sequentially shifting a matching frame 46 having the same size (number of pixels) as the key pattern image 42 in the vertical and horizontal directions in units of one pixel. The The plurality of images to be matched 44 formed are stored in the memory 22a.

  Thereafter, the control device 22 reads the binarized key pattern image 42 and the plurality of matching images 44 from the memory 22a, and correlates the key pattern image 42 and each matching image 44 based on the processing program 22b. Is calculated.

  The correlation (correlation) between the key pattern image 42 and each image to be matched 44 can be calculated by a known method. Specifically, for example, NCC (Normalized Cross-Correlation), SSD (Sum of Squared Difference), SAD (Sum of Absolute Difference), or ZNCC (Zero-means Normalized Cross-Corresponding) may be used.

  When the correlation between the key pattern image 42 and each matched image 44 is calculated as described above, the control device 22 selects the matched image 44 having the strongest correlation as a key pattern. The laser processing apparatus 2 can perform an alignment process or the like based on the key pattern in the workpiece 11b detected in this way.

  As described above, in the key pattern detection method of the present embodiment, a threshold value suitable for binarization processing is set based on the average value of the intensity of each pixel in the image, so that the work reflected in the image Scratches and dirt on the pieces 11a and 11b, grinding marks and the like can be appropriately removed by binarization processing. Thereby, the key pattern can be detected with high accuracy.

  In the present embodiment, since a threshold value suitable for binarization processing is set based on the average value of the intensity of each pixel in the image, a plurality of different threshold values are used in order to investigate the optimum threshold value. There is no need to perform the binarization process once. Therefore, the speed of the key pattern detection method can be increased.

(Embodiment 2)
In the present embodiment, a second aspect of the key pattern detection method will be described. The key pattern detection method according to the second aspect includes a registration step, a workpiece captured image formation step, a first average value calculation step, a second average value calculation step, a threshold setting step, a binarization processing step, a pattern matching Includes steps. The registration step, workpiece captured image forming step, binarization processing step, and pattern matching step are the same as those in the first embodiment.

  In the first average value calculating step, the key pattern image intensity average value is calculated by dividing the total intensity value of each pixel of the key pattern image by the total number of pixels (first pixel number) of the key pattern image. To do.

  In the second average value calculation step, the total value of the shades of each pixel of the workpiece captured image is divided by the total number of pixels of the workpiece captured image (second pixel number), and the workpiece captured image intensity average Calculate the value.

  In the threshold setting step, an average value of the key pattern image density average value and the workpiece captured image density average value is calculated and set as a threshold value for binarization processing (a threshold value for binarization processing). Hereinafter, the key pattern detection method according to the present embodiment will be described in detail.

  In the key pattern detection method according to the present embodiment, first, a registration step and a workpiece captured image formation step are performed. Details of the registration step and the workpiece picked-up image formation step are as described in the first embodiment.

  After the registration step and the workpiece picked-up image formation step, the total value of the shades of each pixel of the key pattern image 42 is divided by the total number of pixels (first pixel number) of the key pattern image 42 to obtain the key pattern image. A first average value calculating step for calculating a lightness average value is performed.

  In the first average value calculating step, first, the control device 22 reads the key pattern image 42 stored in the memory 22a. Based on the processing program 22b, the key pattern image density average value mean_1 is calculated. The calculated key pattern image intensity average value mean_1 is stored in the memory 22a.

The following equation (2) is used to calculate the key pattern image intensity average value mean_1. In Expression (2), n ₁ represents the total number of pixels (first number of pixels) of the key pattern image 42, and K _i represents the intensity of the i-th pixel included in the key pattern image 42.

  After the first average value calculating step, the total value of the shades of each pixel of the workpiece captured image 40b is divided by the total number of pixels (second pixel number) of the workpiece captured image 40b, and the workpiece imaging is performed. A second average value calculating step for calculating an image intensity average value is performed.

  In the second average value calculating step, first, the control device 22 reads the workpiece captured image 40b stored in the memory 22a. Then, based on the processing program 22b, the workpiece captured image density average value mean_2 is calculated. The calculated workpiece captured image intensity average value mean_2 is stored in the memory 22a.

The following formula (3) is used to calculate the workpiece captured image density average value mean_2. In Expression (3), n ₂ represents the total number of pixels (second pixel number) of the workpiece captured image 40b, and W _j represents the intensity of the j-th pixel included in the workpiece captured image 40b. Represent.

  After the first average value calculating step and the second average value calculating step, a threshold value used for the subsequent binarization process based on the key pattern image intensity average value mean_1 and the workpiece captured image intensity average value mean_2. A threshold setting step for setting (binarization processing threshold) is performed.

  In the threshold setting step, first, the control device 22 reads the key pattern image intensity average value mean_1 and the workpiece captured image intensity average value mean_2 stored in the memory 22a. Based on the processing program 22b, an average value mean_3 of the key pattern image density average value mean_1 and the workpiece captured image density average value mean_2 is calculated.

  Thereafter, the calculated average value mean_3 is set as a threshold value used in the subsequent binarization process and stored in the memory 22a. Thus, in the present embodiment, the key pattern image intensity average value mean_1 obtained by dividing the total intensity value of each pixel of the key pattern image 42 by the total number of pixels of the key pattern image 42, and the workpiece captured image 40b. An average value of the total value of the shades of the respective pixels is divided by the total number of pixels of the workpiece captured image 40b and the workpiece captured image density average value mean_2 is set as a threshold value for the binarization process.

  After the threshold setting step, a binarization processing step and a pattern matching step are performed. Details of the binarization processing step and the pattern matching step are as described in the first embodiment.

  As described above, also in the key pattern detection method of the present embodiment, the threshold value suitable for the binarization process is set based on the average value of the intensity of each pixel in the image, so that it is reflected in the image. Scratches and dirt on the workpieces 11a and 11b, grinding marks and the like can be appropriately removed by binarization processing. Thereby, the key pattern can be detected with high accuracy.

  Further, since a threshold value suitable for the binarization process is set based on the average value of the intensity of each pixel in the image, a plurality of binarization processes are performed using different threshold values in order to investigate the optimum threshold value. There is no need to implement. Therefore, the speed of the key pattern detection method can be increased.

  The structures, methods, and the like described in this embodiment can be combined as appropriate with any of the structures, methods, and the like according to the other embodiments.

  In addition, this invention is not limited to description of the said embodiment, A various change can be implemented. For example, in the above-described embodiment, the pattern matching step is performed after the binarization processing step. However, the pattern matching step is not necessarily performed.

  Moreover, in the said embodiment, although illustrated about the aspect which implements a 1st average value calculation step and a 2nd average value calculation step after a registration step and a workpiece picked-up image formation step, the order of each step Is not limited to this.

  At least the first average value calculating step may be performed after the registration step, and the second average value calculating step may be performed after the workpiece captured image forming step. Similarly, the order of the steps constituting the key pattern detection method can be changed and implemented as long as no contradiction occurs.

  Moreover, in the said embodiment, although the workpiece 11a for registration and the workpiece 11b for a process are distinguished, this invention is not limited to this. For example, a part of the processing workpiece 11b may be used as a registration workpiece.

  In addition, the configurations, methods, and the like according to the above-described embodiments can be changed as appropriate without departing from the scope of the object of the present invention.

2 Laser processing device 4 Base 6 Case 8 Processing head 10 Camera 12 Laser irradiation unit 14 Chuck table 16 Cassette elevator 18 Cassette 20 Monitor 22 Control device 22a Memory 22b Processing program 30a, 30b, 30c Captured image 32 Key pattern image 34 Matching Frame 36 Matched image 40a Captured image 40b Workpiece captured image 42 Key pattern image 44 Matched image (pattern)
46 Matching frame 11, 11a, 11b Work piece 13 Device area 15 Peripheral surplus area 17a, 17b Street (scheduled processing line)
19, 19a, 19b Device P pattern Pk Key pattern

Claims (3)

A key pattern detection method for detecting a key pattern on a workpiece,
A registration step of registering a key pattern image having a first pixel number formed by imaging a feature point on the workpiece;
A workpiece imaging image forming step of imaging a workpiece and forming a workpiece imaging image of a second number of pixels having a wider field of view than the key pattern image;
Average brightness obtained by dividing the sum of the shade of each pixel of the key pattern image and the shade of each pixel of the workpiece captured image by the sum of the first pixel number and the second pixel number A threshold setting step for calculating a value and setting it as a threshold for binarization processing;
A key pattern detection method comprising: a binarization processing step for performing binarization processing on the key pattern image with the threshold for binarization processing set in the threshold setting step. A key pattern detection method for detecting a key pattern on a workpiece,
A registration step of registering a key pattern image having a first pixel number formed by imaging a feature point on the workpiece;
A workpiece imaging image forming step of imaging a workpiece and forming a workpiece imaging image of a second number of pixels having a wider field of view than the key pattern image;
A first average value calculation step of calculating a key pattern image intensity average value by dividing the total intensity of each pixel of the key pattern image by the first number of pixels;
A second average value calculating step of dividing the total density of each pixel of the workpiece captured image by the second number of pixels to calculate a workpiece captured image intensity average value;
Calculating an average value of the key pattern image intensity average value calculated in the first average value calculating step and the workpiece captured image intensity average value calculated in the second average value calculating step; A threshold setting step to be a threshold for the optimization
A key pattern detection method comprising: a binarization processing step for performing binarization processing on the key pattern image with the threshold for binarization processing set in the threshold setting step. After performing the binarization processing step, the key pattern obtained by performing binarization processing on the workpiece captured image with the threshold for binarization processing and performing binarization processing on the workpiece captured image The pattern matching step of calculating the correlation by performing pattern matching on the image and detecting the region where the highest correlation is obtained as a key pattern is further provided. Key pattern detection method.

Images