JP2006049868A - Image forming method of component - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image forming method of a component in which the profile of the component is clarified. <P>SOLUTION: The image forming method comprises: a step (S101) of picking up the color image of an electronic component by means of a scanner 170 and forming the color image thereof, and steps (S102, S104) of transforming the color image formed at the color image forming step (S101) into a component monochrome image where the outline of the electronic component is represented by gray on a white or block background, and an electrode monochrome image where the electrode of the electronic component is represented by gray. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an image generation method for generating an image of a component such as an electronic component, and more particularly to an image generation method for generating an image used for specifying the characteristics of a component mounted on a substrate. It is.

  Conventionally, a method of generating an image of an electronic component by imaging the electronic component with a black and white CCD camera has been proposed in order to recognize the external shape, dimensions, positional deviation, and the like of the electronic component mounted on a substrate. (For example, refer to Patent Document 1).

The conventional part image generation method generates an image with contrast between the part and the background by adjusting the color tone of the illumination that illuminates the part.
Japanese Patent Laid-Open No. 10-208017

  However, in the above conventional part image generation method, the generated image is simply a black and white grayscale image captured by a black and white CCD camera, so the contrast between the part and the background, the electrode of the part and other parts, There is a problem that the contrast of the component is insufficient and the form of the component becomes unclear.

  Accordingly, the present invention has been made in view of such a problem, and an object thereof is to provide a component image generation method for clarifying the form of a component.

  In order to achieve the above object, a component image generation method of the present invention is a method for generating an image of the component used for specifying the characteristics of the component mounted on a board, and the component is colored. By imaging, a color image generation step for generating a color image of the component, and the color image generated in the color image generation step is converted into a black and white gray image in which the form of the component is represented by black and white shading. Conversion step. Here, in the color image generation step, the part is imaged in color so that the background becomes a single color, and in the conversion step, the color image generated in the color image generation step is changed to white or black as the background. Thus, the black and white image may be converted into the black and white image in which the outer shape of the component is expressed in black and white.

  As a result, the color image of the component is converted into a black-and-white gray image, and the form of the component in the black-and-white gray image can be clarified by appropriately performing the conversion process. Further, by making the background a single color, it is possible to increase the difference in shading between the outer shape of the component and the background, and to clarify the outer shape of the component in the black and white grayscale image. Furthermore, a widely used scanner can be used to generate a color image of a component, and a personal computer or the like can be used to convert a black and white grayscale image. Therefore, a special imaging device for component mounting is used. The cost of image generation can be reduced. In addition, since a black-and-white gray image is finally generated, it is possible to reduce the processing burden in the process of specifying the feature of the component from the image as compared to specifying from the color image.

  The component image generation method may further include a decompression process step of performing a decompression process to widen a shade width on the black and white grayscale image.

  Thereby, since the width of the shade of the black and white shade image is widened, the difference in shade between the outer shape of the component and the background can be further increased, and the outer shape of the component can be made clearer.

  Further, the expansion processing step may further convert the gray value of each pixel smaller than a predetermined gray value among the pixels included in the black and white gray image to a minimum gray value.

  As a result, even if the grayscale width of the monochrome grayscale image is limited as in 256 gradations, the grayscale value of each image smaller than the predetermined grayscale value is set to the minimum value of the 256 grayscales, and the grayscale value is set. By truncating, the shade width of the effective portion of the monochrome shade image can be widened within the limited range.

  In the color image generation step, the color image is generated by imaging a component so that the background is blue. In the conversion step, the color image is generated so that the background is black compared to the component. It may be characterized by being converted into a black and white grayscale image.

  For example, in the conversion step, the conversion formula of ((R + G) / 2−B) / 2 + K represented by a red gradation value R, a green gradation value G, a blue gradation value B, and a constant K is used. The RGB value representing the color of each pixel included in the color image is converted into a black and white gray value.

  This makes it possible to clarify the outline of a non-blue component in a black and white grayscale image.

  In the color image generation step, the color image is generated by imaging a component so that the background is red. In the conversion step, the color image is converted so that the background is black compared to the component. It may be characterized by being converted into a black and white grayscale image.

  For example, in the conversion step, the conversion formula of ((B + G) / 2−R) / 2 + K represented by a red gradation value R, a green gradation value G, a blue gradation value B, and a constant K is used. The RGB value representing the color of each pixel included in the color image is converted into a black and white gray value.

  This makes it possible to clarify the outer shape of a non-red component in a black and white grayscale image.

  Further, the component is an electronic component, and in the conversion step, the color closer to white among the colors included in the color image generated in the color image generation step is made more prominent by black and white shading. The color image may be converted into an electrode black and white grayscale image in which the electrodes of the component are represented by black and white grayscale.

  For example, in the color image generation step, a Lab color system color image is generated, and in the conversion step, when conversion into the electrode black and white grayscale image is performed, for each pixel included in the color image, The L value representing brightness and the a and b values representing chromaticity are converted into black and white grayscale values using a conversion formula of L × (K− (a2 + b2) 1/2) represented by a constant K. To do.

  Electrodes of electronic components appearing in a color image are achromatic bright colors that are close to white, and such a color image is converted into an electrode black-and-white shade image in which the closer the color to white is, the more the black-and-white shade of the electrode is converted. The outline of the electrode of the electronic component can be clarified in the image.

  The component image generation method may further include a filter processing step for removing noise from the electrode black and white grayscale image.

  As a result, noise is removed from the black-and-white electrode grayscale image, so that the outer shape of the electrode of the electronic component can be made clearer. Further, for a relatively small electrode, the influence of noise can be greatly suppressed, and the characteristics of the electrode can be easily specified.

  The component image generation method may further include an expansion processing step for expanding a width of light and shade with respect to the electrode black and white light and shade image.

  As a result, the width of the shade of the electrode monochrome shade image is widened, so that the difference in shade between the electrode of the electronic component and other parts and the background can be further increased to further clarify the outer shape of the electrode of the electronic component. it can.

  Further, the image generation method of the part further includes a gray value that is larger than the gray value of the pixel corresponding to the electrode among the gray values of each pixel included in the electrode black-and-white gray image converted in the conversion step. It is also possible to include a black and white conversion step for converting to a gray value.

  For example, when the body of the electronic component is white, the density values of the body and the electrode are both high and close to white in the electrode black-and-white density image, and the difference between the density of the body and the electrode becomes small. In addition, since the electrode of the electronic component that appears in the color image is darker than white, when the body of the electronic component is white, the gray value of the pixel corresponding to the body of the electrode black and white gray image corresponds to the electrode It becomes a little larger than the gray value of the pixel. Therefore, even if the body of the electronic component is white, by converting the gray value larger than the gray value of the pixel corresponding to the electrode to the minimum gray value, the body is blackened in the electrode black and white gray image, and the electrode Only can be whitened. Therefore, even if the color of the part other than the electrode of the electronic component is white, only the outer shape of the electrode of the electronic component can be clarified.

  The present invention is also realized as a part teaching data creation method or part automatic teaching apparatus for creating part teaching data using the part image generation method, a program for those methods, and a storage medium for storing the programs. be able to. Furthermore, the present invention can be realized as an image generation apparatus that generates an image by the above-described component image generation method. Furthermore, the present invention can also be realized as a component mounting apparatus including the component automatic teaching apparatus.

  The component image generation method of the present invention has the effect of being able to clarify the form of the component.

  Hereinafter, an automatic component teaching apparatus according to an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a configuration diagram showing a configuration of an automatic component teaching apparatus according to an embodiment of the present invention.
The component automatic teaching apparatus 100 uses the image generation method of an electronic component (hereinafter simply referred to as a component) of the present invention to generate component teaching data indicating the shape and characteristic dimensions of the component. A scanner 170 that captures an image of an image, an operation unit 110 that receives an operation by an operator, a display unit 130 that includes, for example, a liquid crystal display, and the scanner 170 and the display unit 130 are controlled based on an operation that is received by the operation unit 110. And a control unit 120. The control unit 120 includes a peripheral device control unit 121, an input / output processing unit 122, and an image processing unit 123.

2A and 2B are external views showing the external appearance of the scanner 170. FIG.
As shown in FIGS. 2A and 2B, the scanner 170 includes a scanner main body 171, a cover 172, and a cover sheet 173.

  The scanner main body 171 includes a substantially rectangular box-shaped housing 171a, an imaging device and a lighting tool embedded in the housing 171a, and a glass plate 171b attached to the upper surface of the housing 171a.

  The cover 172 is rotatably attached to the housing 171a of the scanner body 171. When the operator installs the component 1 on the scanner 170, the cover 172 is moved away from the glass plate 171b of the scanner main body 171 (open state), and when the operator tries to image the component 1, the cover 172 is Then, the scanner body 171 is brought into a state (closed state) so as to cover the glass plate 171b.

  The cover sheet 173 is attached to the surface of the cover 172 facing the glass plate 171b. The cover sheet 173 is blue or red, and the blue cover sheet 173 and the red cover sheet 173 are selectively used according to the color of the component 1 to be imaged.

  When the operator images the part 1, first, the operator determines whether the color of the part 1 to be imaged is blue or red, and if it is not blue, the cover 172 is opened and the blue cover sheet 173 is opened. The cover 172 is attached. Then, the operator places the component 1 on the glass plate 171b and closes the cover 172. In this state, the scanner 170 illuminates the component 1 installed on the glass plate 171b with the illuminator of the scanner body 171 and captures the component 1 in color while moving the image sensor, so that the component with a blue background is present. 1 color image is generated. On the other hand, if the component 1 to be imaged is not red, the operator attaches the red cover sheet 173 to the cover 172 in the same manner as described above, and installs the component 1 on the glass plate 171b. In such a state, the scanner 170 captures the part 1 in color, thereby generating a color image of the part 1 with a red background. In this way, a color image of the component 1 having a blue or red background is generated.

  The input / output processing unit 122 of the control unit 120 acquires the operation content received by the operation unit 110 and outputs the operation content to the peripheral device control unit 121 and the image processing unit 123. Further, the input / output processing unit 122 acquires a color image generated by the scanner 170 and outputs the color image to the image processing unit 123.

  The peripheral device control unit 121 acquires the operation content received by the operation unit 110 via the input / output processing unit 122, and controls the scanner 170 and the display unit 130 based on the operation content.

  The image processing unit 123 performs predetermined processing on the color image of the part generated by the scanner 170 to generate a black and white gray image in which the outer shape of the part is expressed by black and white shading, and the black and white gray image is displayed. It is displayed on the display unit 130.

  Specifically, the image processing unit 123 includes a monochrome image for component outline (hereinafter referred to as a component monochrome image) for specifying the outline of the component, and a monochrome image for electrode outline for specifying the outline of the electrode of the component. A grayscale image (hereinafter referred to as an electrode black and white image) is generated. The component black-and-white image represents the outer shape of the component with black and white shading, and the electrode black-and-white image represents the outer shape of the electrode of the component with black and white shading. The background is black.

  Further, the image processing unit 123 acquires the operation content received by the operation unit 110 via the input / output processing unit 122 and, based on the operation content, component teaching data indicating the external shape and characteristic dimensions of the component. Determine creation conditions for creation. Then, the image processing unit 123 creates component teaching data from the black and white grayscale image and the creation conditions.

  The creation conditions include, for example, the electrode type of the component, the shape of the ground plane of the electrode (hereinafter referred to as the electrode ground plane shape), and the like. The operator operates the operation unit 110 according to the screen displayed on the display unit 130 based on the control by the control unit 120, and inputs the electrode type and electrode ground plane shape of the part.

  FIG. 3A to FIG. 3C are screen display diagrams illustrating an example of a screen displayed on the display unit 130 in order for the image processing unit 123 to determine a creation condition.

  As shown in FIG. 3A, the image processing unit 123 displays a message a1 such as “Please set information necessary for image input” and an electrode type field for selecting and displaying an electrode type name from a pull-down menu. a2, an electrode type selection button a3 for selecting an electrode type from a figure, a surface shape column a4 for selecting and displaying the name of the electrode ground plane shape from a pull-down menu, and a electrode ground plane shape for selecting from the figure A surface shape selection button a5, a determination button a6 that prompts the user to determine the creation conditions based on the contents displayed in the electrode type field a2 and the surface shape field a4, and a cancel button a7 that prompts to stop the determination of the creation conditions Is displayed on the display unit 130.

  When the screen shown in FIG. 3A is displayed, the operator operates the operation unit 110 to directly display the electrode type with the largest number of parts in the electrode type column a2, or select the electrode type. Displayed indirectly using button a3.

  When displaying indirectly, the operator selects the electrode type selection button a3. As a result of the selection, the display unit 130, as shown in FIG. 3B, for example, a graphic button b1 showing nine electrode type figures, a name column b2 showing the names of the nine electrode types in a list format, A type determination button b3 that prompts to determine the electrode type indicated by the selected graphic button b1 and a type cancel button b4 that prompts to stop the selection of the electrode type using the graphic are displayed. Here, the operator finds a figure indicating the electrode type of the part from the figures represented by the nine figure buttons b1, and operates the operation unit 110 to select the figure button b1 representing the figure. To do. When the graphic button b1 is selected, among the names displayed in the name column b2, the name of the electrode type indicated by the selected graphic button b1 is highlighted. Next, when the operator selects the type determination button b3 by operating the operation unit 110, the display unit 130 displays the screen illustrated in FIG. 3A again. In the screen, the electrode type column a2 displays the name of the electrode type indicated by the graphic button b1 selected as described above. Thus, by using the electrode type selection button a3, the operator can easily input the name from the figure without knowing the name of the electrode type of the part.

  In addition, when the screen shown in FIG. 3A is displayed, the operator operates the operation unit 110 to directly display the electrode ground surface shape of the part in the surface shape column a4, Displayed indirectly using the shape selection button a5.

  When displaying indirectly, the operator selects the surface shape selection button a5. As a result of the selection, as shown in FIG. 3C, the display unit 130 displays, for example, a figure c1 indicating two electrode ground plane shapes, a check box c2 corresponding to each figure c1, and a checked check box c2. A surface shape determination button c3 that prompts the user to determine the electrode grounding surface shape corresponding to 1 and a surface shape cancel button c4 that prompts the user to stop selecting the electrode grounding surface shape using the graphic are displayed. Here, the operator finds a figure indicating the electrode ground plane shape of the part from the two figures c1 and operates the operation unit 110 to check the check box c2 corresponding to the figure (the black dot is set). input). Next, when the operator selects the surface shape determination button c3 by operating the operation unit 110, the display unit 130 displays the screen illustrated in FIG. 3A again. In the screen, the surface shape column a4 displays the name of the electrode ground plane shape corresponding to the check box c2 checked as described above. Thus, by using the surface shape selection button a5, the operator can easily input the name from the figure without knowing the name of the electrode ground surface shape of the component.

  Thus, the operator selects the decision button a6 in the state where the names of the electrode type and the electrode ground plane shape are selected and displayed in the electrode type column a2 and the surface shape column a4 of the screen shown in FIG. Then, the electrode type and electrode ground plane shape of the selected and displayed name are determined as the component teaching data creation conditions.

  The image processing unit 123 specifies an algorithm indicating a part to be dimension-measured, a measurement processing method thereof, and the like from the creation conditions determined in this way and the black and white grayscale image. Then, the image processing unit 123 measures the dimension of the part in the part, that is, the characteristic dimension from the black and white grayscale image using the algorithm, and creates part teaching data indicating the characteristic dimension and the shape of the part. To do. In addition, the image processing unit 123 causes the display unit 130 to display the component teaching data.

  Such component teaching data is used for recognizing misalignment of components in a component mounting apparatus that mounts components on a substrate. That is, when a component mounting apparatus mounts a component on a board, the component is sucked by the head and transported to a predetermined part on the substrate, and the component is mounted on the part. The captured part is imaged with a camera. Then, the component mounting apparatus recognizes the positional deviation of the component from the imaging result and the above-described component teaching data, and drives the head or the like so as to eliminate the positional deviation. Thereby, components can be accurately mounted on a predetermined portion of the board.

FIG. 4 is a flowchart showing the overall operation of the component automatic teaching apparatus 100.
First, the component automatic teaching apparatus 100 determines a creation condition based on an operation by an operator (step S100).

  Next, the component automatic teaching apparatus 100 generates a color image by imaging the component in color (step S101), generates a component black-and-white image from the color image (step S102), and generates an electrode monochrome from the color image. An image is generated (step S104). In other words, in steps S102 and S104, the component automatic teaching apparatus 100 converts the color image of the component into a black and white grayscale image in which the shape of the component is expressed in black and white.

  Then, the automatic component teaching apparatus 100 detects the external shape of the component and the external shape of the electrode of the component from the component black and white image and the electrode black and white image (step S106). Such detection of the outer shape is detected from the difference between the grayscale values of black and white. At this time, the automatic component teaching apparatus 100 identifies a predetermined algorithm with reference to the creation conditions determined in step S100.

  Finally, the component automatic teaching apparatus 100 measures the characteristic dimensions of the component from the detected component outer shape and electrode outer shape using the predetermined algorithm, and generates component teaching data indicating the measurement result. (Step S108).

  Here, a method for generating a component black-and-white image using the chroma key technique in step S102 described above will be described in detail.

FIG. 5 is an explanatory diagram for explaining a method for generating a component black-and-white image according to the present embodiment.
The image processing unit 123 acquires a color image of a component from the scanner 170, and generates a monochrome image of the component with a white component and a black background from a blue or red color image. The image processing unit 123 increases the contrast between the component and the background by performing an expansion process on the component black-and-white image. Here, the component monochrome image generated first from the color image is referred to as an intermediate component monochrome image, and the component monochrome image generated by the decompression process is referred to as a final component monochrome image.

  Specifically, when the image processing unit 123 generates an intermediate component black-and-white image from a color image, for example, for each pixel, an RGB gradation value (RGB value) of 256 gradations, and a monochrome gradation of 256 gradations. Convert to value.

  When recognizing that the background of the color image is blue, the image processing unit 123 uses the blue conversion formula X = ((R + G) / 2−B) / 2 + 128) to calculate the RGB value indicated by each pixel of the color image. It is converted into a black and white grayscale value X of 256 gradations. Here, R represents the gray value of 256 gradations of red, G represents the gradation value of 256 gradations of green, and B represents the gradation value of 256 gradations of blue. Further, (R + G) / 2−B indicates the blueness of the pixel, and indicates a small value if the blueness is strong. In addition, a pixel having a large gray value X is represented so that white is strong, and a pixel having a small gray value X is represented so that black is strong.

  For example, if the pixel included in the color image is yellow, the RGB value of the pixel is (255, 255, 0), and the image processing unit 123 sets the RGB value (255, 255, 0) to X = Conversion into a monochrome gray value of ((255 + 255) / 2-0) / 2 + 128 = 255. Further, if the pixel included in the color image is green, the image processing unit 123 has the RGB value of the pixel as (0, 255, 0), and therefore the RGB value is set to X = ((0 + 255) / 2−. 0) / 2 + 128 = 192 to convert to black and white grayscale value. Similarly, if the pixel included in the color image is white, the image processing unit 123 converts the RGB value (255, 255, 255) of the pixel into a black and white grayscale value of X = 128, and the color If the pixel included in the image is dark blue, the RGB value (0, 0, 160) of the pixel is converted into a black and white gray value of X = 48, and if the pixel included in the color image is blue, The RGB value (0, 0, 255) of the pixel is converted into a black and white gray value of X = 0.

  Further, when the image processing unit 123 recognizes that the background of the color image is red, each pixel of the color image indicates using a red conversion formula (X = ((B + G) / 2−R) / 2 + 128). The RGB value is converted into a black and white grayscale value X of 256 gradations. Here, (B + G) / 2−R indicates the redness of the pixel, and indicates a small value if the redness is strong. Similarly to the above, pixels with a large shading value X are represented so that white is strong, and pixels with a small shading value X are represented so that black is strong.

  In this way, for a non-blue component, an intermediate component monochrome image is generated from the blue background color image by the blue conversion formula, and for a non-red component, a red background image is used. Since the intermediate part black-and-white image is generated by the red conversion formula, the contrast between the part and the background can be increased.

  Next, when performing the expansion process, the image processing unit 123 uses a reference gray value (hereinafter referred to as a reference gray value) that characterizes black and white clarity from the gray value of each pixel included in the intermediate component black and white image. The maximum gray value (hereinafter referred to as the maximum gray value) is searched. Then, the image processing unit 123 expands the width from the reference gray value to the maximum gray value to a width of 0 to 255.

FIG. 6 is an explanatory diagram for explaining an example of the decompression process.
When the image processing unit 123 finds the reference gray value a and the maximum gray value b from the gray values of each pixel included in the intermediate component black-and-white image, the width of the gray values a to b is set to a width of 0 to 255. Stretch to.

  Specifically, the image processing unit 123 sets, for each pixel, a gray value X that is equal to or greater than the reference gray value a before the expansion process of the intermediate component black and white image, and a gray value X1 = 255 (X−a) / (B−a), and the gray value X smaller than the standard gray value a is converted into the gray value X1 = 0 (the minimum value of 256 gradations).

  By such a stretching process, the contrast between the component and the background can be further increased.

Here, the reference gray value a is searched using a histogram of gray values.
FIG. 7 is an explanatory diagram for explaining a method of searching for the reference gray value a.

  The image processing unit 123 creates a histogram representing the number of pixels for each gray value in the intermediate component black and white image. The image processing unit 123 that has created the histogram identifies the gray value c having the largest number of pixels from a group having a high gray value in the histogram, that is, a group having a strong white color. Then, for example, the image processing unit 123 sets 60% of the gray value c as the reference gray value a.

  As described above, in this embodiment, since the width of the shade of the intermediate part black-and-white image is widened, the difference in shade between the outline of the part and the background can be increased to make the outline of the part clearer. Further, among the gray values of each pixel included in the intermediate part black-and-white image, all the gray values smaller than the reference gray value a are set to 0. It is possible to widen the shade width of the effective portion of the intermediate part black-and-white image.

  FIG. 8 is a flowchart showing the operation of the image processing unit 123 generating the final component black-and-white image of the component (the operation in step S102 in FIG. 4).

  First, the image processing unit 123 acquires a color image from the scanner 170 (step S200), and converts the color image into an intermediate component monochrome image (step S202). Next, the image processing unit 123 retrieves the reference gray value a and the maximum gray value b from the intermediate component black-and-white image (step S204), and uses the reference gray value a and the maximum gray value b for the intermediate component black-and-white image. The decompression process is performed (step S206). As a result, a final part black-and-white image of a part with high contrast is generated.

  As described above, in this embodiment, in order to convert a color image of a component whose background is blue or red into a component black-and-white image, the external shape of the component is obtained by using a blue conversion formula or a red conversion formula focusing on the background color. The outline of the part in the part black-and-white image can be clarified by increasing the difference in shading between the background and the background. Furthermore, in the present embodiment, since the scanner 170 is a commercially available scanner, and the control unit 120, the operation unit 110, and the display unit 130 can be configured by a personal computer or the like, the cost of image generation can be suppressed. Further, in the component automatic teaching apparatus 100 of the present embodiment, the outer shape is detected from the final component black-and-white image, and the dimensions of the component are measured from the outer shape. Therefore, compared with the case of detecting the outer shape directly from the color image, The burden on the dimension measurement process can be reduced.

  In this embodiment, the component monochrome image is generated so that the component is white and the background is black. However, the component monochrome image may be generated so that the component is black and the background is white.

  Next, the method for generating an electrode black-and-white image in step S104 described above will be described in detail.

  FIG. 9 is an explanatory diagram for explaining a method for generating an electrode black-and-white image according to the present embodiment.

  The image processing unit 123 acquires a color image of a component from the scanner 170, and converts the color image into an electrode monochrome image by using an electrode enhancement conversion formula. By this electrode emphasis conversion formula, in the electrode black-and-white image, it is possible to make the color close to white in the color image stand out by black and white shading so that the electrode of the component is white and the other portions and the background are black. Then, the image processing unit 123 removes noise by performing a filtering process on the electrode black and white image, and further performs a decompression process to increase the contrast between the electrode of the component and other parts. Here, the electrode black and white image generated first from the color image is referred to as the first intermediate electrode black and white image, and the electrode black and white image generated by the filtering process is referred to as the second intermediate electrode black and white image, which is expanded. The generated electrode monochrome image is referred to as a final electrode monochrome image.

  Specifically, when generating the first intermediate electrode black-and-white image from the color image, the image processing unit 123 first converts the color image represented by the RGB color system into a color image represented by the Lab color system. To do.

FIG. 10 is an explanatory diagram for explaining the RGB color system and the Lab color system.
When the color image output from the scanner 170 is configured in, for example, the BMP (Bit Map) format, the color image is expressed in the RGB color system. That is, each pixel included in the color image has an RGB value composed of a red shade value R, a green shade value G, and a blue shade value B, and is a color specified by the RGB value. Represents.

  Each pixel included in a color image of the Lab color system has a Lab value composed of an L value indicating brightness and an a value and a b value indicating chromaticity, and is a color specified by the Lab value. Represents. Here, the a-axis indicates a color range from red to green, and the b-axis indicates a color range from yellow to blue. That is, the image processing unit 123 converts the RGB value indicated by each pixel of the color image acquired from the scanner 170 into a Lab value.

  Next, the image processing unit 123 uses the electrode enhancement conversion formula X = L × (100− (a2 + b2) 1/2), and for each pixel included in the color image of the Lab color system, the Lab indicated by the pixel is displayed. The value is converted into a black and white gray value X. As a result, a first intermediate electrode monochrome image is generated. The L value takes a value from 0 to 100, the a value and the b value take values from −100 to +100, respectively, and the constant 100 of the electrode emphasis conversion formula is a numerical value that can take a value, b value, etc. It is determined according to the range.

FIG. 11 is a diagram illustrating the distribution of each color included in a general component (electronic component).
Each pixel included in the color image of the Lab color system representing a general part is distributed as shown in FIG.

  That is, since the pixels corresponding to the electrodes of the component are bright achromatic colors, the L value is large and the a value and the b value are distributed in the vicinity of 0. Further, since most of the parts other than the electrode of the part have many dark colors, many pixels corresponding to the part are distributed in the vicinity of a small L value.

  Here, when the image processing unit 123 converts the Lab value of each pixel included in the above-described color image into the gray value X using the electrode emphasis conversion formula, the Lab color system color included in the color image is converted. Each represented pixel is distributed as shown in FIG. In other words, pixels corresponding to electrodes representing a color close to white are distributed in a region having a large gray value X, and pixels corresponding to a package or background representing a dark color such as black are small in gray value X. Many are distributed in the area. Therefore, in the first intermediate electrode black-and-white image generated by such a conversion formula, the electrodes are displayed prominently so that the electrode of the component is white and the other portions and the background are black. That is, the outer shape of the electrode of the component can be clarified.

  When generating the second intermediate electrode monochrome image from the first intermediate electrode monochrome image, the image processing unit 123 performs a filtering process using a median filter on the first intermediate electrode monochrome image. That is, the image processing unit 123 arranges the gray values X of 3 × 3 pixels including the target pixel in the first intermediate electrode monochrome image in ascending order, and sets the gray value X located in the center of the target pixel as the target pixel. A gray value X is assumed. The image processing unit 123 performs such processing with each pixel as a target pixel, thereby generating a second intermediate electrode monochrome image.

  By such filtering processing, it is possible to suppress the influence of noise on the second electrode black-and-white image and make the outer shape of the electrode of the component clearer. In addition, the influence of noise can be greatly suppressed for relatively small electrodes.

  Next, when performing the expansion process, the image processing unit 123 selects a reference gray value (hereinafter referred to as a reference gray value) that characterizes black and white clarity from the gray value of each pixel included in the second intermediate electrode black and white image. ) And the maximum gray value (hereinafter referred to as the maximum gray value). Then, the image processing unit 123 expands the width from the reference gray value to the maximum gray value to a width of 0 to 255. Such decompression processing is performed in the same manner as the decompression processing described with reference to FIGS.

  As described above, in this embodiment, in order to widen the shade width of the second intermediate electrode black and white image, the difference in shade between the outer shape of the component electrode and the other portions and the background is increased, and the outer shape of the electrode of the component is increased. It can be made clearer. Further, among the gray values of each pixel included in the second intermediate electrode black and white image, all the gray values smaller than the reference gray value a are set to 0. Within the limit range, the shade width of the effective portion of the second intermediate electrode monochrome image can be widened.

  FIG. 12 is a flowchart showing the operation of the image processing unit 123 for generating the final electrode monochrome image of the component (the operation in step S104 in FIG. 4).

  First, the image processing unit 123 acquires a color image from the scanner 170 (Step S300), and converts the color system of the color image from the RGB color system to the Lab color system (Step S302). Next, the image processing unit 123 generates a first intermediate electrode black-and-white image by converting the Lab value indicated by each pixel included in the color image into a black and white grayscale value X using an electrode enhancement conversion formula ( Step S304).

  Then, the image processing unit 123 performs a filtering process on the first intermediate electrode black and white image to generate a second intermediate electrode black and white image (step S306), and the reference gray value a from the second intermediate electrode black and white image. And the maximum gray value b is searched (step S308). Next, the image processing unit 123 performs a decompression process on the second intermediate electrode black and white image using the reference gray value a and the maximum gray value b, thereby generating a final electrode black and white image (step S310).

  As described above, in the present embodiment, the electrode of the component that appears in the color image is an achromatic bright color that is close to white, and such a color image is converted into an electrode black-and-white image in which the color close to white is more conspicuous. Therefore, the outer shape of the electrode of the component can be clarified in the electrode black and white image. Furthermore, in the present embodiment, since the scanner 170 is a commercially available scanner, and the control unit 120, the operation unit 110, and the display unit 130 can be configured by a personal computer or the like, the cost of image generation can be suppressed. In the automatic component teaching apparatus 100 according to the present embodiment, the outer shape is detected from the final electrode black-and-white image, and the dimensions of the electrode of the component are measured from the outer shape. In comparison, the burden on the dimension measurement process can be reduced.

  In this embodiment, the component black-and-white image is generated so that the electrode of the component is white and the other portions are black. However, the black and white electrode is generated so that the electrode of the component is black and the other portions are white. An image may be generated.

  Further, in the generation of the electrode black-and-white image in the present embodiment, the color of the background of the color image of the component need not be close to white, and may be blue, red, green, or the like.

  In addition, in the generation of the electrode monochrome image in the present embodiment, the electrode monochrome image is generated so that the closer to the white color included in the color image, the whiter the electrode monochrome image is. However, the extremely white portion included in the electrode monochrome image is further generated. May be blackened. For example, when a color image of the component 1 having a white body is converted into an electrode monochrome image using the above-described electrode enhancement conversion formula, not only the electrode of the component 1 but also the body becomes white in the electrode monochrome image. It may be difficult to detect the outer shape of the. Therefore, using the fact that the actual electrode color (silver) is darker than white, the gray value of each pixel included in the electrode monochrome image is larger than the gray value of the pixel corresponding to the electrode. The gray value is converted to the minimum gray value (X = 0). That is, pixels that are whiter than the pixels corresponding to the electrodes are blackened.

  For example, in an electrode black-and-white image, if the gray value of a pixel corresponding to a silver electrode is 240 and the gray value of a pixel corresponding to a white body is 255, the outer shape of the electrode is determined from the electrode black-and-white image. It is difficult to detect. However, as described above, when the gray value of 250 or more among the gray values of each pixel included in the electrode black and white image is converted to 0, the gray value of the pixel corresponding to the electrode in the converted electrode black and white image is 240. On the other hand, since the gray value of the pixel corresponding to the body is 0, only the electrode is emphasized, and the outer shape of the electrode can be easily detected from the electrode monochrome image.

  Such an additional conversion process is performed when the body of the component 1 is white, and in order to determine whether or not the body is white, a black and white grayscale image (generated separately for detecting the outer shape of the component 1) ( Component black and white image) is used. That is, the average gray value of the part corresponding to the body of the component 1 in the black and white gray image for the part outline is measured, and when the gray value is larger than the threshold value, the body is white. Determined. In addition, although the part corresponding to a body changes with types of the components 1, it is a center part of the components 1 generally.

  Alternatively, a pixel having a gray value of 250 or more may be searched from the entire electrode black and white image, and the gray value of the pixel may be converted to 0, or may be searched from the region corresponding to the component 1 of the electrode black and white image, The gray value of the pixel may be converted to 0.

(Modification)
Here, a modified example of the automatic component teaching apparatus in the present embodiment will be described.

  The component automatic teaching apparatus according to this modification is provided in a component mounting apparatus that mounts a component on a board.

  FIG. 13 is an external view showing an external appearance of a component mounting apparatus including an automatic component teaching apparatus according to this modification.

  The component mounting apparatus 500 includes a guide rail 514 for transporting the substrate 2, a parts feeder 530 that supplies components such as resistors and capacitors, and ICs such as SOP (Small Outline Package) and QFP (Quad Flat Package) ( A part tray 532 on which a relatively large part such as an integrated circuit) or a connector is placed; a head 528 that sucks and mounts the part supplied from the part feeder 530 or the part placed on the part tray 532 onto the substrate 2; The automatic component teaching apparatus 300 includes a recognition unit 540 that recognizes the positional deviation of the component sucked by the head 528 by referring to the component teaching data created by the automatic component teaching apparatus 300.

  In such a component mounting apparatus 500, component teaching data of a component attracted by the head 528 is created by the component automatic teaching apparatus 300 in advance of component mounting, and when the component is mounted, the component mounting apparatus 500 is attracted by the head 528. The component is picked up, and the positional deviation of the component adsorbed by the head 528 is recognized using the image pickup result and the component teaching data prepared in advance.

  When the substrate 2 is mounted on the guide rail 514 and is transported from the loader unit 516 of the component mounting apparatus 500 to the mounting position 518 by driving a transport belt (not shown), and the component is mounted at the mounting position 518, It is conveyed to the unloader unit 520.

  The head 528 includes a plurality of suction nozzles 534 that suck components at the tip, and moves in a direction along the plane of the substrate 2 by driving a motor (not shown). By such movement of the head 528, parts are conveyed. The head 528 conveys the component to an imaging position to be described later when creating the component teaching data. On the other hand, at the time of component mounting, the head 528 first transports the component to the imaging position, images the component at the imaging position, and then transports the component to the board 2 at the mounting position 518. The head 528 drives the suction nozzle 534 to mount the component on the substrate 2.

  The component automatic teaching apparatus 300 according to this modification is provided so as to be able to image a component at an imaging position. That is, the component automatic teaching apparatus 300 images the component sucked by the suction nozzle 534 of the head 528 from the side opposite to the head 528 (from the bottom in FIG. 13).

FIG. 14 is a configuration diagram showing a configuration of an automatic component teaching apparatus 300 according to this modification.
The component automatic teaching apparatus 300 according to this modification includes a display unit 330, an operation unit 310, and a control unit 320. Furthermore, instead of the scanner 170 of the component automatic teaching device 100 of the above embodiment, a head 528 is provided. An imaging unit 370 that captures an image of the component 1 that is adsorbed to the component 1 is provided.

  The display unit 330 and the operation unit 310 according to this modification have the same functions as the display unit 130 and the operation unit 110 of the above-described embodiment. The operation unit 310 is disposed in the component mounting apparatus 500 so that the operator can operate it (see FIG. 13), and the display unit 330 is disposed near the operation unit 310.

  The imaging unit 370 includes a camera 372 that captures images in color, a lighting device 371 that applies white light to the component 1 that is attracted to the head 528, and a background plate 373 that is attached to each suction nozzle 534 of the head 528. Yes. The surface on the component 1 side of the background plate 373 is blue or red. If the component 1 sucked by the suction nozzle 534 is not blue, the blue background plate 373 is used for the suction nozzle 534, and if the component 1 sucked by the suction nozzle 534 is not red, the red background plate 373 is used. A background plate 373 is used for the suction nozzle 534.

  Such an image pickup unit 370 takes the component 1 with the background plate 373 as the background when the component 1 attracted to the head 528 is conveyed to the image pickup position A when creating the component teaching data and mounting the component. An image is taken and a color image of the component 1 with a blue or red background is generated.

  The control unit 320 according to this modification includes a peripheral device control unit 321, an input / output processing unit 322, and an image processing unit 323. The peripheral device control unit 321 and the input / output processing unit 322 have the same functions as the peripheral device control unit 121 and the input / output processing unit 122 of the above embodiment. The image processing unit 323 basically has the same function as the image processing unit 123 of the above embodiment, but further determines a component size as a component teaching data creation condition.

  FIG. 15 is a screen display diagram illustrating an example of a screen displayed on the display unit 330 in order for the image processing unit 323 to determine a creation condition.

  The image processing unit 323 causes the display unit 330 to display a size column a21 for selecting and displaying a component size from a pull-down menu. When the operator operates the operation unit 310 to display the dimension of the maximum side of the part 1 as the part size in the size column a21 and selects the determination button a6, the part size is determined as a creation condition. Further, the image processing unit 323 according to the present modification displays the screen illustrated in FIG. 15 on the display unit 330 before the imaging of the component 1 by the imaging unit 370, and the electrode type, the electrode ground plane shape, and the component size. Is previously determined as a creation condition.

  In the component automatic teaching apparatus 300 according to this modification example, when the component 1 sucked by the suction nozzle 534 of the head 528 is conveyed to the imaging position A, the camera 372 of the imaging unit 370 is illuminated by the illumination tool 371. The part 1 thus taken is imaged. At this time, the camera 372 adjusts the angle of view based on the component size determined by the image processing unit 323 and images the component 1 at the angle of view. Then, the camera 372 outputs a color image of the component 1 as the imaging result to the control unit 320. When the blue background plate 373 is attached to the head 528, a color image with a blue background is output, and when the red background plate 373 is attached to the head 528, the background is red. A color image is output. The component automatic teaching apparatus 300 according to the present modification generates a black and white grayscale image from the color image thus captured in the same manner as in the above embodiment, and the component teaching data based on the black and white grayscale image and the creation conditions. Create

  The recognition unit 540 provided in the component mounting apparatus 500 is configured by, for example, a CPU (Central Processing Unit) that controls each device provided in the component mounting apparatus 500, for example.

  The recognition unit 540 controls the illuminator 371 and the camera 372 to cause the camera 372 to image the component 1. Then, the recognizing unit 540 applies the suction nozzle 534 of the head 528 to the suction nozzle 534 of the head 528 based on the component teaching data created in advance by the component automatic teaching apparatus 300 and the color image or black-and-white gray image generated by the camera 372 when mounting the component. Recognize the displacement of the sucked component 1. Specifically, the recognizing unit 540 determines the displacement in the direction (horizontal direction) along the suction surface of the component 1 sucked by the suction nozzle 534 and the shift in the rotation direction of the component 1 on the suction surface as described above. Recognize as a gap. Next, the recognition unit 540 drives the head 528 and the suction nozzle 534 so as to eliminate the positional deviation. As a result, the component 1 is accurately mounted at a predetermined site on the board 2.

  Since the component mounting apparatus 500 has both a function for generating component teaching data and a function for mounting components, compared to a system that includes the component automatic teaching apparatus 100 and the component mounting apparatus, respectively. The position shift can be recognized more accurately. That is, in the system described above, the scanner 170 that captures the component 1 of the component automatic teaching apparatus 100 is different from the camera that captures the component 1 adsorbed by the head of the component mounting apparatus. Since the camera 372 used to create the image and the camera 372 that images the component 1 adsorbed by the adsorption nozzle 534 are the same, it is possible to prevent the occurrence of an error due to the difference in the imaging means.

  In the modified example, the component mounting apparatus 500 creates component teaching data in advance of component mounting, and uses the component teaching data and the image generated by the camera 372 to shift the position of the component 1. Although recognized, the image generation processing of the above embodiment is also applied to capturing an image when recognizing the positional deviation. That is, when the camera 372 picks up the component 1 conveyed to the image pickup position A and generates a color image at the time of mounting the component, the image processing unit 323 of the control unit 320 at that time causes the component 1 from the color image. A black and white gray image is generated. Then, the recognizing unit 540 recognizes the position shift of the component 1 attracted to the head 528 during component mounting by detecting the outer shape of the component 1 from the black-and-white gray image or measuring a characteristic dimension. . Even in such a case, the same effect as described above can be obtained.

  Further, the image processing unit 323 may have the function of the recognition unit 540.

  The part image generation method of the present invention has an effect that the form of the part can be clarified. For example, the part automatic teaching that teaches the operator the characteristic dimension of the electronic part from the image of the electronic part, for example. The present invention can be applied to a device and a component mounting device that recognizes a positional shift of a component based on the teaching result.

It is a block diagram which shows the structure of the components automatic teaching apparatus in embodiment of this invention. (A) And (b) is an external view which shows the external appearance of a scanner same as the above. (A)-(c) is a screen display figure which shows an example of the screen which an image processing part same as the above displays on a display part in order to determine creation conditions. It is a flowchart which shows the whole operation | movement of a component automatic teaching apparatus same as the above. It is explanatory drawing for demonstrating the production | generation method of a component monochrome image same as the above. It is explanatory drawing for demonstrating an example of the expansion | extension process same as the above. It is explanatory drawing for demonstrating the method to search a reference | standard light / dark value same as the above. It is a flowchart which shows the operation | movement which the image processing part same as the above produces | generates the final component monochrome image of components. It is explanatory drawing for demonstrating the production | generation method of an electrode monochrome image same as the above. It is explanatory drawing for demonstrating the RGB color system and Lab color system same as the above. It is a figure which shows distribution of each color contained in a general component (electronic component). It is a flowchart which shows the operation | movement which the image process part in embodiment of this invention produces | generates the final electrode monochrome image of components. It is an external view which shows the external appearance of the component mounting apparatus provided with the component automatic teaching apparatus which concerns on the modification same as the above. It is a block diagram which shows the structure of the components automatic teaching apparatus which concerns on a modification same as the above. It is a screen display figure which shows an example of the screen which an image processing part which concerns on a modification same as the above displays on a display part in order to determine creation conditions.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Parts automatic teaching apparatus 110 Operation part 120 Control part 121 Peripheral equipment control part 122 Input / output processing part 123 Image processing part 130 Display part 170 Scanner

Claims (20)

A method of generating an image of the component used to identify the characteristics of the component mounted on a substrate,
A color image generation step of generating a color image of the component by imaging the component in color;
A component image generation method comprising: a conversion step of converting the color image generated in the color image generation step into a black and white grayscale image in which the shape of the component is represented by black and white shades. In the color image generation step,
Image the part in color so that the background is a single color,
In the conversion step,
The color image generated in the color image generation step is converted into the black and white grayscale image in which the background is white or black and the external shape of the component is expressed by black and white shades. 2. A method for generating an image of a component according to item 1. The component image generation method further includes:
The method for generating an image of a component according to claim 2, further comprising an expansion process step of performing an expansion process for expanding the width of light and shade on the black and white image. In the extension processing step,
4. The component image generation method according to claim 3, wherein a gray value of each pixel smaller than a predetermined gray value among pixels included in the black and white gray image is converted into a minimum gray value. In the color image generation step,
Generate the color image by imaging the component so that the background is blue,
In the conversion step,
3. The component image generation method according to claim 2, wherein the color image is converted into a black and white grayscale image so that the background becomes black compared to the component. In the conversion step,
Each pixel included in the color image using a conversion formula of ((R + G) / 2−B) / 2 + K represented by a red gradation value R, a green gradation value G, a blue gradation value B, and a constant K 6. The method of generating an image of a component according to claim 5, wherein the RGB value representing the color of the component is converted into a grayscale value of black and white. In the color image generation step,
Generate the color image by imaging the component so that the background is red,
In the conversion step,
3. The component image generation method according to claim 2, wherein the color image is converted into a black and white grayscale image so that the background becomes black compared to the component. In the conversion step,
Each pixel included in the color image using a conversion formula of ((B + G) / 2−R) / 2 + K represented by a red gradation value R, a green gradation value G, a blue gradation value B, and a constant K The component image generation method according to claim 7, wherein the RGB value representing the color of the component is converted into a grayscale value of black and white. The component is an electronic component,
In the conversion step,
By making the color closer to white out of the colors included in the color image generated in the color image generation step stand out with black and white shading, the color image is made up of black and white electrodes in which the electrodes of the component are represented by black and white shading. The component image generation method according to claim 2, wherein the image is converted into a grayscale image. The component image generation method further includes:
The component image generation method according to claim 9, further comprising a filter processing step for removing noise from the electrode black-and-white grayscale image. The component image generation method further includes:
The component image generation method according to claim 10, further comprising an expansion process step of expanding a width of light and shade with respect to the electrode black and white gray image. In the color image generation step,
Generate a color image of the Lab color system,
In the conversion step,
When performing conversion to the electrode black and white image,
For each pixel included in the color image, an L value representing the brightness of the pixel and an a value and a b value representing chromaticity are expressed by L × (K− (a2 + b2) 1 / The component image generation method according to claim 9, wherein the image is converted into black and white grayscale values by the conversion formula of 2). The component image generation method further includes:
A black and white conversion step of converting a gray value larger than the gray value of the pixel corresponding to the electrode among the gray values of each pixel included in the electrode black and white image converted in the conversion step into a minimum gray value; 10. The component image generation method according to claim 9, wherein: A method for creating component teaching data indicating characteristics of a component mounted on a board,
A color image generation step of generating a color image of the component by imaging the component in color;
A conversion step of converting the color image generated in the color image generation step into a black and white grayscale image in which the form of the component is represented by black and white shading; and identifying the feature of the component from the black and white grayscale image; A method for creating part teaching data, comprising: a data creating step for creating part teaching data to be shown. An image generation device for generating an image of the component used for specifying the characteristics of the component mounted on a substrate,
Color image generation means for generating a color image of the component by capturing the component in color so that the background is a single color;
Conversion means for converting the color image generated by the color image generation means into a black-and-white gray image in which the background is white or black and the outer shape of the component is expressed in black and white. Image generation device for parts. The component image generation apparatus further includes:
The component image generation apparatus according to claim 15, further comprising an expansion processing unit configured to perform an expansion process for expanding a width of light and shade for the black and white image. A component automatic teaching apparatus for creating component teaching data indicating characteristics of a component mounted on a board,
An image generating apparatus for parts according to claim 15 or 16,
A component automatic teaching device comprising: a data creating unit that identifies a feature of the component from a monochrome grayscale image generated by the image generating device and creates component teaching data indicating the feature. A component mounting apparatus for mounting a component on a board,
Color image generating means for generating a color image of the component by imaging the component in color;
Conversion means for converting the color image generated by the color image generation means into a black-and-white shading image in which the form of the component is expressed by black and white shading; and identifying the feature of the component from the black-and-white shading image; Data creation means for creating the component teaching data shown;
By holding and moving the component, a head for mounting the component on a predetermined part on the substrate;
A component mounting apparatus comprising: recognition means for recognizing a positional deviation of the component held by the head using the component teaching data created by the data creating device. A component mounting apparatus for mounting a component on a board,
A head for mounting the component on a predetermined part on the substrate by holding and moving the component;
Color image generation means for generating a color image of the component by capturing the component held by the head in color so that the background is a single color;
Conversion means for converting the color image generated by the color image generation means into a black-and-white gray image in which the background is white or black and the outer shape of the component is expressed in black and white;
A component mounting apparatus comprising: a recognizing unit that recognizes a position shift of a component held by the head using the black and white grayscale image converted by the converting unit.   The program which makes a computer perform the step contained in the image generation method of the components of any one of Claims 1-13.

Images