JP4119824B2 - Method and apparatus for processing inspection of photovoltaic device panel - Google Patents

Description

  The present invention relates to a photovoltaic device panel processing inspection method and apparatus, and more particularly to a method and apparatus for inspecting the processing state of grooves formed in each film by laser etching.

Conventionally, a conductive transparent electrode film, an amorphous conductor film, and a metal electrode film are laminated on a transparent substrate, and a groove is formed by irradiating each film with a laser beam and separated from each other on the transparent substrate. There is known a photovoltaic device such as a solar battery configured by configuring a plurality of cells and connecting the cells in series.
In the laser etching processing of grooves, normally, continuous laser beams having a circular pattern are continuously or intermittently moved while irradiating a laser beam having a circular pattern, thereby forming continuous grooves in the form of overlapping circular patterns. To be done.
As a method for increasing the processing speed without increasing the output of the laser, a laser beam having a substantially elliptical pattern is moved in the major axis direction (see, for example, Patent Document 1).
And in order to manage the processing quality of the photovoltaic device panel manufactured in this way, it is necessary to inspect the shape of the groove formed in each film by laser etching.

Since the groove formed in the photovoltaic device panel has an extremely fine groove width of several tens of μm, it must be inspected while being enlarged using a microscope or the like.
Conventionally, the quality of the processed state of the groove has been judged by a skilled engineer observing an image magnified by a microscope.
As a general technique, a method of displaying image data measured with a scanning laser microscope as a three-dimensional image as a method for managing the shape of an etched portion of a thin film metal material has been proposed (see, for example, Patent Document 2). .)
JP 2002-33495 A (paragraph [0013], FIG. 1 etc.) Japanese Patent Laid-Open No. 10-158864 (paragraph [0007], FIG. 5, FIG. 6 etc.)

  However, since the grooves formed by laser etching are extremely fine as described above and have a complicated shape, even a skilled engineer can determine the quality of the processing state. Took a long time. In addition, if burrs are formed on each film by laser etching, it may cause contact with or short circuit with the transparent electrode film, resulting in a problem that the photoelectric conversion efficiency is significantly reduced. However, it is unrealistic for a human to search for a burr of about several μm over the entire wide panel, and there is an inconvenience that much inspection time is required even if possible.

  In addition, the groove shape when processed by a laser beam with a substantially elliptical pattern has a new parameter of flatness due to the difference between the major axis and the minor axis. Compared with the case of, it becomes remarkably difficult. For this reason, there is an inconvenience that an individual difference that the determination result is different for each engineer to be inspected, and the inspection time is further prolonged.

  Further, the contour of the groove formed by the laser etching process also includes a portion that is not clear due to gradation, and the judgment is also different depending on individual differences and measurement conditions, so that the accuracy of the pass / fail judgment is lowered. It was a problem.

  The present invention has been made in view of the above circumstances, and it is intended to provide a photovoltaic device panel processing inspection method and apparatus capable of accurately inspecting in a short time and enabling on-line inspection and consequently 100% inspection. It is aimed.

In order to solve the above problems, the present invention provides the following means.
That is, the present invention is a photovoltaic device comprising a conductive transparent electrode film, an amorphous semiconductor film, and a metal electrode film laminated on a transparent substrate, each of which is irradiated with a pulsed laser beam to form a groove. A processing inspection method for inspecting a processing state of a groove in a power device panel, the step of imaging the groove from the transparent substrate side, and extracting features including the shape of the groove by performing image processing on the captured image And a step of determining a processing state of the groove based on the extracted features , wherein the groove penetrates only the conductive transparent electrode film and reaches the transparent substrate. An amorphous semiconductor film groove that penetrates only the amorphous semiconductor film and reaches the conductive transparent electrode film; and the conductive transparent electrode film that penetrates the amorphous semiconductor film and the metal electrode film With a groove reaching up to For example, the light and performing said only to the amorphous semiconductor film and the groove through said metal electrode layer reaching to the conductive transparent electrode film, a burr analysis for determining the presence or absence of burrs A method for processing and inspecting an electromotive force device panel is provided.
According to the present invention, since the captured image is image-processed to extract the features including the groove shape, it is possible to accurately determine the machining state of the groove in a short time without depending on the skill of the inspection engineer. it can.

In the above invention, it is preferable to determine whether or not the feature is a burr formed by processing, and the feature extraction step is a burr based on the number of pixels of the image protruding from the outer edge image of the groove.
Further, in the above invention, the feature is a burr formed by processing, and it is determined whether or not the feature extraction step is a burr based on the number of pixels of an image protruding from a tangent line in contact with the outer edge image of the groove. It is good as well.
According to these inventions, it is possible to accurately determine the presence or absence of burrs by eliminating an unrealistic inspection work in which an inspection engineer visually searches for fine burrs over the entire surface of a wide photovoltaic device panel. it can. As a result, a short circuit between the films due to burrs can be prevented in advance, and a decrease in photoelectric conversion efficiency can be prevented.

In the above invention, the groove may be formed by overlapping a plurality of substantially elliptical laser etching patterns, and the feature may be a contour shape of each substantially elliptical pattern.
According to the present invention, even in a substantially elliptical laser etching pattern, it is possible to apply the groove shape with high accuracy by calculation processing such as fitting to a similar elliptical shape or calculating by an elliptic equation without depending on the skill of an inspection engineer. It is possible to determine the quality of the machining state by grasping the contour shape.

Furthermore, in the above invention, the groove is formed by penetrating the metal electrode film and the amorphous semiconductor film to the conductive transparent electrode film by overlapping a plurality of substantially elliptical laser etching patterns. The feature is a shape of a boundary line between the amorphous semiconductor film and the conductive transparent electrode film formed inside each substantially elliptical pattern, and the feature extraction step is a captured image. The boundary line may be extracted by binarizing.
According to the present invention, since the captured image is binarized by a preset threshold value, gradation is generated regardless of the skill or subjectivity of the inspection engineer and regardless of changes in measurement conditions. It is possible to accurately extract a boundary line of a simple contour shape and determine whether the machining state is good or bad.

In addition, the present invention provides a photovoltaic device comprising a conductive transparent electrode film, an amorphous semiconductor film, and a metal electrode film laminated on a transparent substrate, each of which is irradiated with a pulsed laser beam to form a groove. In a power device panel, a processing inspection device that inspects a processing state of a groove, including a camera that images the groove from the transparent substrate side, and includes the shape of the groove by performing image processing on an image captured by the camera. An image processing unit that extracts a feature and a processing state determination unit that determines a processing state of a groove based on the extracted feature , wherein the groove penetrates only the conductive transparent electrode film and the transparent substrate. A conductive transparent electrode film groove reaching up to, an amorphous semiconductor film groove reaching only the amorphous transparent film through the amorphous semiconductor film, and penetrating through the amorphous semiconductor film and the metal electrode film To the conductive transparent electrode film In a groove reaching said to the groove only reaching the amorphous semiconductor film and the conductive transparent electrode film through said metal electrode layer, to perform a burr analysis for determining the presence or absence of burrs A photovoltaic device panel processing inspection apparatus is provided.
According to the present invention, the contour shape of the groove formed in each film is imaged through the transparent substrate by the operation of the camera. The captured image is subjected to image processing by the operation of the image processing unit to extract features, and the processing state determination unit is used to determine the processing state of the groove based on the features. As a result, it is possible to quickly and accurately determine the quality of the machining state.

In the above-described invention, a transport unit that transports the photovoltaic device panel, and a camera that moves the camera in a direction along the surface of the transparent substrate with respect to the photovoltaic device panel transported by the transport unit It is preferable to provide a moving means.
According to the present invention, the photovoltaic device panel conveyed by the operation of the conveying means is inspected while moving the camera along the surface of the transparent substrate. Thus, it is possible to determine the quality of the processing state of the photovoltaic device panel. As a result, 100% inspection can be performed and the reliability of the product can be improved.

  According to the present invention, it is possible to quickly and accurately inspect the processing state of the groove having a complicated shape formed in each film of the photovoltaic device panel, so that online inspection or 100% inspection is possible. As a result, the reliability of the product can be improved.

DESCRIPTION OF EMBODIMENTS Embodiments of a photovoltaic device panel processing inspection method and apparatus according to the present invention will be described below with reference to the drawings.
Before describing each embodiment of the present invention, the structure of a photovoltaic device panel to be inspected will be described below with reference to FIGS.

  As shown in FIG. 1, the photovoltaic device panel 1 includes a conductive transparent electrode film 3 made of a material such as tin oxide on the surface of a transparent substrate 2 made of glass and an amorphous semiconductor made of a material such as silicon. The film 4 and the metal electrode film 5 made of a material such as aluminum are stacked in this order. Each of the films 3 to 5 is formed with a plurality of grooves for separating each film into a plurality. Specifically, the conductive transparent electrode film 3 adjacent to the transparent substrate 2 is formed with a groove 6 that penetrates only the conductive transparent electrode film 3 and reaches the transparent substrate 2. Further, the amorphous semiconductor film 4 adjacent to the conductive transparent electrode film 3 is formed with a groove 7 that penetrates only the amorphous semiconductor film 4 and reaches the conductive transparent electrode film 3. Further, the amorphous semiconductor film 4 and the metal electrode film 5 are formed with grooves 8 that penetrate the amorphous semiconductor film 4 and the metal electrode film 5 and reach the conductive transparent electrode film 3.

In order to manufacture the photovoltaic device panel 1, first, after forming the conductive transparent electrode film 3 on the transparent substrate 2, the groove 6 is formed in the conductive transparent electrode film 3, and the groove 6 is formed. A part of the amorphous semiconductor film 4 is filled in the groove 6 by forming the amorphous semiconductor film 4 on the conductive transparent electrode film 3 formed. Next, a groove 7 is formed in the amorphous semiconductor film 4, and a metal electrode film 5 is formed on the amorphous semiconductor film 4 in which the groove 7 is formed. Then, a groove 8 is formed so as to penetrate the amorphous semiconductor film 4 and the metal electrode film 5 thus laminated at the same time.
The conductive transparent electrode film 3 is formed by, for example, a sputtering method or a thermal CVD method. The amorphous semiconductor film 4 is formed by, for example, a plasma CVD method. Further, the metal electrode film 5 is formed by, for example, a sputtering method.

  The conductive transparent electrode film 3 is separated by the material of the groove 6 and the amorphous semiconductor film 4 filled in the groove 6, and amorphous by the material of the groove 7 and the metal electrode film 5 filled in the groove 7. The semiconductor film 4 is separated, and the amorphous semiconductor film 4 and the metal electrode film 5 are separated by the groove 8. Thus, the photovoltaic device panel 1 is formed in a state where a plurality of cells are separated on the transparent substrate 2 and adjacent cells are connected in series by the metal electrode film 5 filled in the groove 7. Is manufactured.

Since all these grooves 6 to 8 are formed so as to reach either the transparent substrate 2 or the conductive transparent electrode film 3, when viewed from the transparent substrate 2 side, as shown in FIG. Their contour shape can be grasped.
The grooves 6 and 7 formed in the conductive transparent electrode film 3 and the amorphous semiconductor film 4 are irradiated with a pulsed laser beam while continuously moving in one direction, so that each of the grooves is almost smooth. It is formed in a straight groove shape.
Further, the groove 8 formed through the amorphous semiconductor film 4 and the metal electrode film 5 is uneven by irradiating a laser beam having an approximately elliptical pattern intermittently moving in one direction. It is formed in a continuous linear groove shape.

  In addition, the photovoltaic panel 1 is formed with an insulating groove 9 around the area where the photoelectric effect region and the non-photoelectric region are separated from each other. As shown in FIG. 3, the insulating groove 9 is narrow in the conductive transparent electrode film 3 by irradiating a pulsed laser beam in a state where all the films 3 to 5 are formed in a laminated state. The groove 9a having a groove width, the amorphous semiconductor film 4 and the metal electrode film 5 are formed in two steps including a groove 9b having a wide groove width. Since the insulating groove 9 is also formed so as to reach the transparent substrate 2, the contour shape can be observed from the transparent substrate 2 side as shown in FIG. 4.

[First Embodiment]
Next, a processing inspection apparatus and method for the photovoltaic device panel 1 according to the present embodiment will be described below with reference to FIGS.
As shown in FIG. 5, the processing inspection apparatus 10 according to the present embodiment has a conveyor 11 that carries the manufactured photovoltaic device panel 1 and conveys it, and is disposed opposite to the upper surface of the conveyor 11. A microscope observing device 12 including a camera for observing the photovoltaic panel 1 mounted on the conveyor 11 from the transparent substrate 2 side, and the microscope observing device 12 are moved in two horizontal axes, thereby moving the photovoltaic panel 1 An XY stage 13 that enables observation over the entire surface, an image processing device (image processing unit, processing state determination unit) 14 that processes an image signal captured by the microscope observation device 12, and a result of processing by the image processing device 14 An output monitor 15 and alarm means 16 for notifying the fact based on the quality of the processed state as a result of the image processing are provided. The image processing device 14 also has a function of outputting a position command signal to the XY stage 13 and controlling the camera position.

The image processing device 14 processes the image signal transmitted from the camera, and extracts and extracts features including the shapes of the grooves 6 to 8 formed in the photovoltaic device panel 1 from the image signal. Based on the characteristic, the processing state of the grooves 6 to 8 is determined.
Specifically, the image processing apparatus 14 performs the processing inspection method shown in the flowchart of FIG. In the following description, when the direction of an image is shown, the left and right direction will be described as the X direction and the up and down direction as the Y direction.

That is, the image processing apparatus 14 determines whether or not there is an image in step S1, and proceeds to step S2 if an image signal is input, and ends the process if no image signal is input.
In step S2, the type of image signal is selected. The image signal includes three grooves 6 formed in the conductive transparent electrode film 3, grooves 7 formed in the amorphous semiconductor film 4, grooves 8 formed in the amorphous semiconductor film 4, and the metal electrode film 5. The first image signal in which the grooves 6 to 8 are simultaneously imaged, the second image in which the two grooves 6 and 7 formed in the conductive transparent electrode film 3 and the amorphous semiconductor film 4 are simultaneously imaged There are three types of image signals, a third image signal in which only the signal and the insulating groove 9 are imaged.

  When the input image signal is the first image signal, first, the groove 8 formed in the amorphous semiconductor film 4 and the metal electrode film 5 is analyzed (step S3), and then the amorphous signal is detected. The groove 7 formed in the semiconductor film 4 is analyzed (step S4), and the groove 6 formed in the conductive transparent electrode film 3 is analyzed (step S5). Further, it is analyzed whether or not burrs are formed in the groove 8 (step S6), and the analysis result is displayed and stored (step S7).

First, step S3 will be described.
The analysis of the grooves 8 formed in the amorphous semiconductor film 4 and the metal electrode film 5 includes a Yamaya difference analysis (step S31), an inner ellipse analysis (step S32), an outer ellipse analysis (step S33), and an ellipse pattern matching. (Step S34) is included.

As shown in FIG. 7, the Yamatani difference analysis S31 firstly performs an initial setting (step S310), and then sets an image (for example, an R image of RGB images) captured by the camera to a predetermined threshold. Binarization is performed according to the value (step S311). FIG. 8 is an image of the groove 8 obtained by binarization.
Second, the binarized image is scanned from bottom to top along the Y direction to determine the presence or absence of the binarized image, and the lower outline is traced. By repeating this process in the X direction, the lower outer shape image of the groove 8 is extracted as shown in FIG. 9 (step S312). In the extraction process, unnecessary images that are separated from the lower outer shape are removed.

Thirdly, the obtained lower outer shape image is scanned in the X direction to detect the position of each mountain valley and determine the distance Δ in the Y direction between adjacent mountain valleys (step S313).
The obtained distance Δ is recorded as a processed dot unevenness (parameter R0).
Fourth, if there are, for example, four or more places where the calculated distance Δ is equal to or greater than a predetermined value, the process is terminated, and if it is less than four places, the process proceeds to detection of the upper outline image. (Step S314). The reason why the number is four is to secure the number of data necessary for the subsequent ellipse analysis and may be any number.

The detection of the upper outer shape is performed by repeating the process of determining the presence or absence of the binarized image from the lower outer shape position upward in the Y direction in the X direction. (Step S315).
Next, the obtained upper outer shape image is scanned in the X direction to detect the position of each mountain valley and determine the distance Δ between adjacent mountain valleys (step S316).
Similarly to the above, when there are, for example, four or more places where the calculated distance Δ is equal to or larger than a predetermined specified value, the process is terminated, and when there are less than four places, the flag F = 1 is set. (Step S318) and the process ends (Step S317).
Note that step S35 is flag F = 1 when neither the lower outline image nor the upper outline image has four or more predetermined valley differences as a result of the above-described peak / valley difference analysis S31. This is a step of detecting and detecting the end of the process.

The inner ellipse analysis S32 is a step of analyzing an elliptical white region (reference numeral 17 in FIG. 8) formed near the center of each elliptical pattern constituting the image of the groove 8, and the groove 8 is electrically conductive and transparent. It is grasped as a region reaching the electrode film 3.
First, as shown in FIG. 10, the white region 17 binarizes an image captured by the camera using a predetermined threshold value (step S321). This step S321 may be shared with step S311.
Secondly, a white area 17 having a predetermined area existing in the obtained binarized image, for example, a white area 17 having 150 pixels or more and 3000 pixels or less is detected (step S322).
Thirdly, if the number of white areas 17 detected as described above is less than four, the analysis is terminated assuming that there is no connection of processed dots (S323: parameter R6).
Fourth, when there are four or more white regions 17 having a predetermined area, the center position of each white region 17 is obtained, and as shown in FIG. 11, the lowest outer shape image extracted in step S31 is obtained. The white region 17 'having the center position above 75% of the width (maximum width: MAX) between the lower end and the uppermost end of the upper outer shape image is excluded (step S324).
Then, for each white area 17 thus obtained, as shown in FIG. 12, the center coordinates, the starting point coordinates of the circumscribed rectangle 18, the area of the white area, the width of the white area (parameter R8), the white area Each information such as the height (parameter R9) is acquired (step S325).

The outer ellipse analysis S33 performs the following processing on the outer contour line constituting the outer shape of the groove 8 at a position corresponding to the white region 17 detected in the inner ellipse analysis step S32 (FIG. 13).
First, as shown in FIG. 14, the valley position is detected by scanning the left side of the center coordinates of the white region 17 with respect to the detected outline, and the lower side of the center coordinates is scanned. Detect the mountain position. Thereby, the peaks and valleys of the outer contour region are detected (step S331).
Next, an ellipse 19 centering on the central coordinates and passing through the mountain position and the valley position is obtained (step S332).
Then, using the obtained ellipse 19, as shown in FIG. 15, the start point coordinates, height (parameter R 1) and length of the rectangle 20 circumscribing the ellipse 19, and the valleys and valleys of the outer contour shape Distance between the two (parameter R2), the area of the outer ellipse 19, the area ratio of the outer ellipse 19 and the inner ellipse 17 (parameter R7), the short axis side length of the outer ellipse 19, the long axis side length of the outer ellipse 19, Each information such as cost (parameter R3) is acquired (step S333).

  In addition, regarding the overlap margin of the outer ellipse 19, in addition to a method (parameter R3 (1)) obtained by calculating 1/2 of the overlap of the ellipse 19 drawn as described above, each pixel on the outer contour line The major axis of the ellipse 19 is calculated using the ellipse equation based on the coordinate position, and half of the overlap calculated from the relationship between the central coordinate and the major axis of the adjacent ellipse 19 is calculated ( Either parameter R3 (2)) or a method (parameter R3 (3)) obtained as ½ of the width of the black area formed between adjacent white areas 17 after binarization may be used.

The pattern matching S34 stores an image including an ellipse pattern at an arbitrary position from the detected outer contour shape as a model image, and between the model image and an image arranged in another region. This is done by calculating the similarity.
Similarity can be obtained by a method that can express the degree of similarity, such as a method that obtains the area of the mismatched part by overlapping the model image and the target image as a known method, or a method that obtains the normalized correlation coefficient. . In the former, since the recognition rate is low due to slight deformation of the figure and various noises, the latter method using the normalized correlation coefficient is often used. In this case, the similarity between the position and the reference model can be known by searching for the position where the normalized correlation coefficient is maximized. (Parameter R5).

Next, step S4 will be described.
The R image obtained by the camera is subjected to spatial filter processing to search for the groove contour shape. A portion where the image continues without being largely interrupted is regarded as the contour line of the groove 7, and when two contour lines are searched, the interval between both contour lines is calculated (parameter T1). If the interval dimension is within a predetermined range, the position is stored. The threshold value is updated and the above operation is repeated, and it is assumed that the groove 7 exists at the position where the most detection is performed. When the image is largely interrupted, it is determined that the groove 7 is not continuous (parameter T2).
In step S5, the same process as in step S4 is performed (parameters S1, S2).
If the grooves 7 and 6 are not detected in the above processing, the same processing is performed after switching to the G image.

Next, step S6 will be described.
Step S6 includes steps S61 to S66 as shown in FIG.
First, the R image obtained by the camera is binarized with a predetermined threshold value (step S61).
Second, the obtained binarized image is processed to fill the white area 17 (step S62). As a result, as shown in FIG. 17, a binarized image is obtained in which the entire region sandwiched between the outer contour lines is filled with black.

Third, an upper trace (step S63) and a lower trace (step S64) are performed on the obtained binarized image.
The upper trace S63 scans the upper contour line of the binarized image from the outside of the black region 21, thereby detecting the upper outer trace S631 that detects the outer contour shape and the upper contour line. , And an upper inner trace S632 for detecting the outer contour shape by scanning from the inside of the black region 21. When the shape of the actual upper contour line 22 having burrs is the shape shown in FIG. 18A, the upper outer trace S631 repeats the process of scanning downward in the Y direction in the X direction. The detected outer contour shape 23 is as shown in FIG. Further, the upper inner trace S632 becomes 18 (c) upward in the Y direction.
Similarly, also in the lower trace S64, the lower outer trace S641 and the lower inner trace S642 are performed on the lower outline of the binarized image.

Fourth, as a result of the upper inner trace S632 obtained as described above, the presence / absence of a location A that is separated by 3 pixels or more in the vertical direction is inspected (step S651). If there is a location A that is 3 pixels or more away, the result of the upper outer trace S631 is searched within a range of ± 20 pixels from that position. It is checked whether or not it exists (step S652). If there is a location B that is 3 pixels or more away, as shown in FIG. 18D, the location B is detected by the outer contour shape 23 detected by the upper outer trace S631 and the upper inner trace S632. A distance C from the outer contour shape 24 is calculated (step S653). Then, it is inspected whether or not the distance C is 5 pixels or more (step S654), and if it is away, it is determined that a burr exists (step S655).
The same applies to the lower burr detection S66.
The number of detected burrs is integrated for those having a size of 10 μm or more, and stored as parameter R4.

Further, the distance between the three grooves 6 to 8 is also calculated and stored as parameters A1 and A2.
The results of the various image processing performed in this way are displayed on the monitor 15 and stored in an appropriate storage medium, and the image processing device 14 determines whether the processing state will be described later. become.

  Next, when the image signal input in step 2 is the second image signal, the same processing as steps S4 to S6 (steps S8 to S10) is performed. Since the processing content is the same, it is omitted.

Next, when the image signal input in step 2 is the third image signal, the shape analysis of the insulating groove 9 is performed.
First, a part of the background image is extracted from the R image signal in the third image signal, and the average luminance is calculated (step S11).

  Second, outer edge processing (step S12) and inner edge processing (step S13) are performed. The outer edge shows the groove wall contour of a portion 9 b that penetrates the amorphous semiconductor film 4 and the metal electrode film 5 and reaches the conductive transparent electrode film 3 in the two-stage insulating grooves 9. The inner edge is a groove wall contour of a portion 9 a that is disposed in the vicinity of the center in the width direction of the two-stage insulating groove 9 and reaches the transparent substrate 2 through the conductive transparent electrode film 3.

As shown in FIG. 19, the outer edge process S12 includes an edge detection process S121 and an edge tangent detection process S122.
As shown in FIG. 20, the edge detection S121 scans the R image signal of the third image signal downward in the Y direction from the background image side, detects a portion where the luminance changes greatly, and repeats it in the X direction. Is done. The edge detection S121 is performed on the image from both the upper and lower directions in the Y direction. Thereby, the outer edge 25 is detected.

  For example, as shown in FIG. 20, the edge tangent detection S122 integrates the number of pixels in which the outer edge exists in the X direction, and sets the position where the ratio to the total number of pixels in the X direction is 20% or more as the outer edge. It is determined that the position is the tangent 26 position on the mountain side. Similarly, a position where the above ratio is 10% or more is searched upward from a position moved about 15 pixels downward from the tangential position on the mountain side in the Y direction, and the position is the position of tangent 27 on the valley side of the outer edge. Judge. Edge tangent detection S122 is also performed on the image from both the upper and lower directions in the Y direction. The distance between the two mountain-side edge tangents 26 is calculated and stored as parameter I1.

The inner edge process S13 also includes an edge detection process S131 and an edge tangent detection process S132, as shown in FIG.
In the edge detection process S131, as shown in FIG. 22, the luminance of each pixel arranged inside the outer edge 25 is inspected, and the pixel 31 having the lowest luminance detected inside the pixel 30 having the highest luminance is detected. Is recognized as the inner edge. The inner edge processing S13 is also performed from both the upper and lower directions of the image signal.
The edge tangent detection S132 is performed in the same manner as the edge tangent detection S122 for the outer edge 25. Similarly, the distance between the two edge tangents 28 on the valley side is calculated and stored as the parameter I2.

Further, a difference in distance between the outer edge tangent 26 on the mountain side and the inner edge tangent 28 on the valley side, which is arranged above and below in the Y direction, is also calculated and stored as the parameter I3 (see FIG. 4).
Also, the center line O of the two edge tangents 28 of the inner edge is calculated, and the center from the edge tangent 26 of any outer edge with respect to the width dimension W between the peak tangents 26 of the two outer edges is calculated. A ratio d / W with the distance d to the line O is calculated and stored as a degree of symmetry (parameter I7).

  Third, the connection determination of the insulating groove is performed (step S14). In connection determination S14, when the outer edge 25 is detected, if there are, for example, 10 or more places where the edge cannot be detected, it is determined that the insulating groove 9 is cut. (Parameter I5). Alternatively, when the luminance is inspected in the X direction in the insulating groove 9 and pixels having luminance equal to the average luminance of the background image (for example, average luminance −10 or more) continue, the insulating groove 9 is cut. You may judge.

  Fourth, it is determined whether or not burrs are generated in the insulating groove 9 (step S15). The burr determination S15 determines, for example, that there is a possibility that a burr has occurred when there is a place where each pixel of the outer edge image is separated from the adjacent pixel by 3 pixels or more in the Y direction. Are accumulated (parameter I4 (1)). Alternatively, the number of pixels that the outer edge image protrudes from the outer edge tangent line 25 may be integrated, and if more than a predetermined number of pixels protrude, it may be determined that burrs have occurred and the number may be integrated (parameter I4 ( 2)).

Fifth, the similarity of the shape of the outer edge 25 arranged above and below the Y direction of the insulating groove 9 is calculated (step S16).
Specifically, first, an R image is binarized with a predetermined threshold value among images obtained by imaging with a camera (step S161). Next, a model image is created by cutting out an image including one of the upper and lower outer edges over a predetermined range (step S162). Then, the model image is inverted by a straight line along the X direction, and the obtained inverted model image is compared with an image including the other outer edge by pattern matching.
The method for obtaining the similarity is the same as that for the parameter R5 (parameter I6).

Sixth, the edge image surrounded by the peak-side edge tangent 26 and the valley-side edge tangent 27 detected in step S122 is binarized to form a white area 29 as shown in FIG. The 29 Y-direction height and area are calculated. This is performed for the upper and lower outer edges in the Y direction, and the area of each white region 29 is calculated (step S17). In addition, the difference in the area of the white region 29 (parameter I8) and the difference in height (parameter I9) are calculated.
Then, these calculated various values are displayed on the monitor 15 and recorded on an appropriate recording medium (step S18).

  For each parameter, a pass standard value may be set in advance, and the calculated value obtained as described above may be displayed on the monitor together with the pass standard value. Then, if the calculated value is out of the acceptable reference value, it may be displayed on the monitor (for example, displayed in red) or notified to the operator by alarm means.

  Moreover, you may decide to alert | report the cause which exceeded the acceptance standard value by the kind of parameter. For example, when burrs are indicated by parameters R4 and I4, or when the groove width dimensions of parameters T1, T2, I1, R1, etc. are poor, film quality improvement is often necessary, and an alarm to that effect is reported. You may decide to do it. If the other parameters deviate from the acceptance standard value, there may be a problem with the processing equipment, and if a warning to that effect is to be issued, maintenance of the processing equipment should be performed at an early stage. Is effective.

As described above, according to the photovoltaic device panel processing inspection method and apparatus according to the present embodiment, since the captured image is subjected to image processing to extract features including the shapes of the grooves 6 to 9, the inspection is performed. There is an effect that the processing state of the groove can be accurately determined in a short time without depending on the skill of an engineer.
In particular, the processing state of the groove is grasped by the number of burrs, and there is no unrealistic inspection work in which an inspection engineer visually searches for fine burrs over the entire surface of a wide photovoltaic device panel. It can be determined whether or not. As a result, there is an effect that a short circuit between the films due to burrs can be prevented in advance and a decrease in photoelectric conversion efficiency can be prevented.

Further, according to the photovoltaic device panel processing inspection method and apparatus according to the present embodiment, even when adopting a substantially elliptical laser etching pattern to increase the processing speed while suppressing the output, the inspection engineer's Without depending on skill, it is possible to determine the quality of the machining state by accurately grasping the contour shape of the groove 8 by calculation processing such as fitting to a similar elliptic shape or calculating by an elliptic equation.
In particular, the shape of the boundary line between the amorphous semiconductor film and the conductive transparent electrode film formed inside each substantially elliptical pattern of the groove 8 processed by overlapping the substantially elliptical laser etching pattern is Even in the case of gradation, the captured image is binarized by a preset threshold value, so that the boundary line can be set regardless of the skill or subjectivity of the inspection engineer and regardless of changes in measurement conditions. It is possible to accurately extract and determine whether the machining state is good or bad.

By adopting a processing inspection apparatus that performs such a processing inspection method, it is possible to determine whether the processing state of the photovoltaic device panel is acceptable or not by incorporating it into a photovoltaic device manufacturing line. As a result, all the inspections can be performed, and the reliability of the product can be improved.
In addition, this invention is not limited to the said embodiment, It cannot be overemphasized that other arbitrary characteristics may be employ | adopted as a characteristic which shows the processing state of a groove | channel.

It is a longitudinal cross-sectional view which shows an example of the photovoltaic apparatus panel inspected in the processing inspection method concerning this invention. It is a top view which shows the photovoltaic apparatus panel of FIG. It is a longitudinal cross-sectional view which shows the cross section of the insulating groove vicinity of the photovoltaic apparatus panel of FIG. It is a top view which shows the photovoltaic apparatus panel of FIG. It is a block diagram which shows the process inspection apparatus of the photovoltaic apparatus panel which concerns on one Embodiment of this invention. It is a flowchart which shows the processing inspection method of the photovoltaic apparatus panel which concerns on one Embodiment of this invention. It is a flowchart which shows the mountain valley difference analysis routine in the method of FIG. It is a figure which shows the binarized image of the groove | channel formed through the amorphous semiconductor film and the metal electrode film. It is a figure which shows the lower side external shape image of the image of FIG. It is a flowchart which shows the inner side ellipse analysis routine in the method of FIG. It is a schematic diagram explaining the center shift | offset | difference white area | region exclusion block of FIG. It is a schematic diagram explaining the various information acquisition block of FIG. It is a flowchart which shows the outer side ellipse analysis routine in the method of FIG. It is a figure explaining the mountain valley detection block of FIG. It is a figure explaining the ellipse calculation block of FIG. It is a flowchart which shows the burr | flash analysis routine in the method of FIG. It is a schematic diagram explaining the white area | region painting block of FIG. FIG. 17 is a schematic diagram for explaining the upper trace block of FIG. 16, wherein (a) is an actual contour shape, (b) is an upper outer trace, (c) is an upper inner trace, and (d) is an upper outer trace and an upper inner trace. It is a figure explaining the space | interval. It is a flowchart which shows the outer edge detection routine in the method of FIG. It is a figure which shows the outer edge detected by the routine of FIG. It is a flowchart which shows the inner edge detection routine in the method of FIG. It is a schematic diagram explaining the inner edge detection routine of FIG. It is a figure which shows the binarized image of the edge image enclosed by the mountain side edge tangent and valley side edge tangent for demonstrating the outer edge area block of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Photovoltaic device panel 2 Transparent substrate 3 Conductive transparent electrode film 4 Amorphous semiconductor film 5 Metal electrode film 6,7,8 Groove 9 Insulation groove (groove)
10 Processing inspection device 11 Conveyor (conveyance means)
12 Microscope observation device (camera)
13 XY stage (camera moving means)
14 Image processing device (image processing unit, processing state determination unit)
25 outer edge

Claims (7)

In a photovoltaic device panel in which a groove is formed by irradiating a pulsed laser beam on each of a conductive transparent electrode film, an amorphous semiconductor film, and a metal electrode film laminated on a transparent substrate. A processing inspection method for inspecting the processing state of
Imaging the groove from the transparent substrate side;
Extracting a feature including the shape of the groove by performing image processing on the captured image;
Determining the machining state of the groove based on the extracted features ,
The groove includes a conductive transparent electrode film groove that penetrates only the conductive transparent electrode film and reaches the transparent substrate, and an amorphous film that penetrates only the amorphous semiconductor film and reaches the conductive transparent electrode film. A semiconductor film groove, and a groove reaching the conductive transparent electrode film through the amorphous semiconductor film and the metal electrode film,
A burr analysis is performed to determine whether or not there is a burr only for the groove that penetrates through the amorphous semiconductor film and the metal electrode film and reaches the conductive transparent electrode film. Device panel processing inspection method. The feature is a burr formed by processing,
The photovoltaic device panel processing inspection method according to claim 1, wherein the feature extraction step determines whether or not the burr is based on the number of pixels of the image protruding from the outer edge image of the groove. The feature is a burr formed by processing,
2. The processing of a photovoltaic device panel according to claim 1, wherein the feature extraction step determines whether or not there is a burr based on the number of pixels of an image protruding from a tangent line that contacts the outer edge image of the groove. Inspection method. The groove is formed by overlapping a plurality of substantially elliptical laser etching patterns,
The method for processing and inspecting a photovoltaic device panel according to claim 1, wherein the characteristic is a contour shape of each substantially elliptical pattern. The groove is formed by penetrating the metal electrode film and the amorphous semiconductor film to the conductive transparent electrode film by overlapping a plurality of substantially elliptical laser etching patterns,
The feature is a shape of a boundary line between the amorphous semiconductor film and the conductive transparent electrode film formed inside each substantially elliptical pattern,
The photovoltaic device panel processing inspection method according to claim 1, wherein the feature extraction step binarizes the captured image and extracts the boundary line. In a photovoltaic device panel in which a groove is formed by irradiating a pulsed laser beam on each of a conductive transparent electrode film, an amorphous semiconductor film, and a metal electrode film laminated on a transparent substrate. A processing inspection apparatus for inspecting the processing state of
A camera that images the groove from the transparent substrate side;
An image processing unit that extracts a feature including the shape of the groove by performing image processing on an image captured by the camera;
A machining state determination unit that determines the machining state of the groove based on the extracted features ;
The groove includes a conductive transparent electrode film groove that penetrates only the conductive transparent electrode film and reaches the transparent substrate, and an amorphous film that penetrates only the amorphous semiconductor film and reaches the conductive transparent electrode film. A semiconductor film groove, and a groove reaching the conductive transparent electrode film through the amorphous semiconductor film and the metal electrode film,
A burr analysis is performed to determine whether or not there is a burr only for the groove that penetrates through the amorphous semiconductor film and the metal electrode film and reaches the conductive transparent electrode film. Equipment panel processing inspection equipment. Conveying means for conveying the photovoltaic device panel;
The photovoltaic device according to claim 6, further comprising: a camera moving unit that moves the camera in a direction along the surface of the transparent substrate with respect to the photovoltaic device panel conveyed by the conveying unit. Panel processing inspection equipment.

Images