JP2008292398A - Method of inspecting floating defect of component electrode and substrate inspecting device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To inspect floating defect of a component electrode by using a color image produced by a color highlight-type optical system similar to that used in inspection of soldering. <P>SOLUTION: A color image of a region on which the inspection of floating defect of the component electrode 20 is necessary, among the inspection regions determined by every soldering parts in which red, green and blue colors are distributed in the color image, is converted into a grayscale image, and the presence or absence of the floating defect is discriminated by extracting an edge from the image after the conversion. In an image converting processing, color parameters of red, green and blue colors are multiplied by brightness converting coefficients α, β, γ set to convert the color image of the soldered part into the grayscale image of small dispersion of density, and a grayscale is decided on the basis of the sum of multiplication values. As a result, the edge 50 of the component electrode 20 is extracted, but the extraction of the edge at the boundary between original color regions of the soldered part can be prevented in the edge extraction processing to the grayscale image after the conversion. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  This invention is a method for inspecting the state of a soldering part on a substrate on which a component mounting board after soldering processing is to be inspected, and inspecting whether or not a floating defect has occurred in a component electrode of a predetermined component. And a substrate inspection apparatus to which this method is applied.

  The applicant has previously developed a number of substrate appearance inspection apparatuses called “color highlight methods”. This color highlight inspection apparatus includes a two-dimensional color camera (hereinafter simply referred to as “camera”) and illumination that irradiates three types of color light of red, green, and blue from different elevation angles. A color image in which the inclined state of the solder inclined surface is expressed by the distribution state of three colors corresponding to each illumination color is generated by performing imaging under illumination by the illumination device (patent) Reference 1).

Japanese Patent Publication No.6-1173

  In the illuminating device described in Patent Document 1, red light is emitted from the direction having the largest elevation angle when viewed from the substrate surface, blue light is emitted from the direction having the smallest elevation angle, and green light is emitted from the direction between them. The Therefore, in the soldered part in the color image, red appears in the part near the flat surface, blue appears in the steep part, and the part has an inclination angle between the red part and the blue part. Green appears. Therefore, the inclination state of the solder surface can be discriminated by the color appearing in the color image.

  Color highlight inspection equipment uses this principle to binarize the image in the inspection area set in the soldered part of the color image, and the areas where the colors red, green, and blue appear To extract. Then, the area or position of each extracted color region is measured, and the obtained measurement value is compared with a pre-registered determination reference value to determine whether the solder inclination state is correct. Hereinafter, this inspection is referred to as “solder inspection”.

Among various components mounted on the substrate, there is a component called “mini-mold component”.
FIG. 7 shows a comparison between a good mounting state and a poor mounting state of the mini-mold component. In the figure, S is a substrate, 20 is an electrode of a mini-mold component, 21 is a component body, and 22 is solder for joining the electrode 20 to the substrate S.

  In the mini-mold component, as shown in FIG. 7 (1), the state in which the component electrode 20 is buried in the solder 22 is a good mounting state, and as shown in FIG. The state exposed from the solder 22 becomes defective. This type of failure is generally called “floating failure”.

  In order to detect such a floating defect as described above, in general, imaging with a monochrome camera is performed under illumination with monochromatic illumination light to generate a grayscale image of a region to be inspected, and the edge of the component electrode 20 is extracted from this image. To do. When the component electrode 20 is exposed on the solder 22, the number of edge pixels extracted exceeds a predetermined threshold value, so that it is determined that there is a floating defect.

  However, the most important inspection for a substrate after soldering is a solder inspection, and the number of components to be subjected to a component electrode floating defect inspection is limited. In view of the usefulness of color highlighting illumination in solder inspection and the efficiency of inspection, it is desirable to use a color highlighting inspection apparatus so that floating inspection can be performed along with solder inspection.

  However, since the component electrode of the mini-mold component has high specular reflectivity, when the image is taken by the color highlight optical system, the color light corresponding to the inclination of the electrode is reflected in the camera out of the light regularly reflected by the electrode. Incident light may be misidentified as a solder surface. In particular, as shown by dotted line frames A and B in FIG. 6, the solder 22 of the mini-mold component has a gently inclined surface (a red color in the image) near the component main body 21, and floats. In the case where a defect occurs, the electrode 20 often has an inclination close to this gently inclined surface, so that it becomes difficult to identify the defect.

  FIG. 8 (1) schematically shows an image generated when an image of a component in which a floating defect is generated by a color highlight optical system. In this figure, for ease of reference, the components corresponding to those shown in FIG. 7 are indicated by the same reference numerals, and the red, green, and blue color regions in the solder 22 are indicated by halftone dot patterns, respectively. . As shown in this figure, in the color image of the soldered portion of the mini-mold component, a green region appears on the outer periphery of the red region showing the gentle inclined surface, and a blue region appears on the outer periphery.

  In the example of FIG. 8A, the upper right electrode 20 of the component main body 21 is exposed from the solder 22, and the image (the hatched portion in the figure) is also close to the red region on the solder 22 side. Colors appear. As described above, when red appears in the same position as when there is no floating defect even when there is a floating defect, it becomes very difficult to distinguish the floating defect.

  Therefore, the inventor studied a method of converting the color image into a gray image and extracting the edge of the component electrode from the converted gray image. However, when the color image of the soldering part generated by the color highlight optical system is converted into a grayscale image, the color difference in the color image is converted as the density difference. For this reason, when edge extraction processing is performed, as shown in FIG. 8B, in addition to the edges 50, 51, and 52 of the component electrode 20, the component main body 21, and the solder 22, edges 53, which indicate boundaries between the color regions, 54 is extracted. Since these edges 53 and 54 are extracted even in a soldering portion where no floating defect has occurred, there is a possibility that the edges 53 and 54 may be mistaken as the edge 50 of the electrode.

  The present invention pays attention to the above-mentioned problems, and when the color image of the soldering portion generated by the color highlight optical system is converted into a grayscale image, the difference in color appearing in the solder portion is reflected as the density difference. An object of the present invention is to perform an inspection of a component electrode floating defect using the same color image used for solder inspection by performing an image conversion process that is not performed.

  In the method of inspecting a component electrode floating defect according to the present invention, an illumination device that irradiates light of different colors from a plurality of directions with different elevation angles with respect to the substrate to be inspected, and regular reflection light from the substrate with respect to illumination light from the illumination device An image pickup device for a color image arranged at a position where light can be incident, and the component mounting board after soldering is picked up by the image pickup device under illumination by the lighting device, and each illumination color in the generated image is obtained. The inspection is performed using a substrate inspection apparatus that inspects a soldering site on the substrate based on a corresponding color distribution state. In this method, a function for calculating a sum of each multiplication value after multiplying each of a plurality of color parameters constituting color data of each pixel by a predetermined coefficient, and an inspection set for a color image to be processed A function of converting the color image in the inspection region into a grayscale image by replacing the color data of each pixel in the region with the value obtained by the calculation is set in the substrate inspection apparatus. Prior to the inspection, for the color image of the soldered part, the representative value of each color parameter in the area where the color appears for each color corresponding to each illumination color is obtained, and each illumination value is obtained using these representative values. A process of calculating each coefficient is performed assuming that the calculation result obtained when the calculation is executed for each color corresponding to the color has the same value.

  In the inspection, a step of setting an inspection area for inspecting a component electrode floating defect at a soldered portion in a color image to be inspected, a value of each color parameter constituting color data of the pixel for each pixel in the inspection area, and Performing the calculation using the values of the coefficients obtained before the inspection, converting the color image in the inspection region into a gray image based on the calculation result, and extracting the edge pixels from the gray image after the conversion When the number of edge pixels exceeding a predetermined threshold is extracted, it is determined that there is a component electrode floating defect in the inspection region.

  In this method, when the luminance data of one unit of color image is subjected to luminance conversion, each color parameter is weighted by a predetermined coefficient, and then the sum of these is obtained. These coefficients are adjusted so that almost the same calculation result can be obtained regardless of the color that appears in the color image of the soldered part. The area corresponding to the portion can be converted into a grayscale image with small variations in density regardless of the original color.

  On the other hand, when the component electrode is exposed and this electrode appears in the image with a color corresponding to a predetermined illumination color, the intensity of the specular reflection light from the component electrode and the intensity of the specular reflection light from the solder Since the difference is reflected in the calculation result, there is a high possibility that the density difference between the component electrode and the solder is converted into a large image.

  As described above, according to the calculation using the above coefficient, the density variation of the portion corresponding to the solder is reduced, and a grayscale image is generated in which the portion corresponding to the component electrode is represented by a density different from that of the solder portion. can do. Therefore, even if edge extraction is performed on the converted image, it is possible to prevent the boundary between the color regions of the solder portion from being erroneously extracted as an edge, and to accurately determine the presence or absence of the component electrode.

  In the above method, the processing for calculating each coefficient before the inspection is performed by using a color image obtained when the imaging device images a non-defective substrate having the same configuration as the substrate to be inspected, and the component to be inspected for defective floating of the component electrode It is desirable to carry out for each soldering part corresponding to. The intensity of the illumination light for the soldered part varies slightly depending on the positional relationship between the two, or the inclination angle of the solder also varies depending on the height and size of the part. This is because it is considered to vary depending on the type. As described above, if the coefficient is calculated for each soldering part that needs to be inspected for floating defects using the color image of the non-defective board, the tint of the solder is the corresponding part of the non-defective board in the color image to be inspected. Therefore, it is possible to convert the color image of the solder portion where each color is distributed into a gray image with extremely small density variation.

  The substrate inspection apparatus to which the above method is applied includes an illumination device that emits light of different colors from a plurality of directions with different elevation angles with respect to the substrate to be inspected, and regular reflection light from the substrate with respect to illumination light from the illumination device. An image pickup device for a color image arranged at a position where incidence is possible, and a color corresponding to each illumination color in the generated image obtained by picking up an image of the soldered substrate under illumination by the illumination device. The soldering site on the substrate is inspected on the basis of the distribution state, and the inspection result is output. This apparatus includes a calculation means for executing a calculation for calculating a sum of each multiplication value after multiplying a plurality of color parameters constituting color data of each pixel by a predetermined coefficient, and a color image of a soldered portion. In addition, when the representative value of each color parameter in the region where the color appears for each color corresponding to each illumination color is obtained and the above calculation is executed for each color corresponding to each illumination color using these representative values Coefficient calculation means for executing processing for calculating each coefficient on the assumption that the obtained calculation results have the same value, and component electrodes among the inspection areas set in each soldering region in the color image generated by the imaging device For each pixel in the inspection area to be inspected for the floating defect, the color parameter value constituting the color data of the pixel and the coefficient calculating means The color image in the inspection area is shaded by the calculation by the calculation means using the value of each coefficient calculated before the inspection and the process of replacing the color data of the pixel by the value obtained by the calculation. Image conversion means for converting to an image, determination means for extracting edge pixels from the grayscale image generated by the image conversion means, and determining the presence or absence of a floating defect of the component electrode based on the extraction result, and a determination result by the determination means Output means for outputting. Further, the discrimination means compares the number of extracted edge pixels with a predetermined threshold value, and when an edge pixel exceeding the threshold value is extracted, the discrimination electrode has a component electrode in the inspection area corresponding to the grayscale image. It is determined that there is a floating defect.

  According to the board inspection apparatus having the above-described configuration, the solder inspection for each soldering portion is performed using the color image generated by the color highlight optical system, and the same as that set for the solder inspection as appropriate. By using the color image in the inspection area, it is possible to perform an inspection for a floating defect of the component electrode.

  In a preferred aspect of the above substrate inspection apparatus, the coefficient calculation means performs the process of obtaining each coefficient using a color image when the imaging apparatus images a non-defective substrate having the same configuration as the substrate to be inspected. This is executed for each inspection area to be processed. The image processing apparatus further includes a registration unit that stores the coefficient calculated by the coefficient calculation unit for each inspection region to be processed by the image conversion unit.

  According to the above aspect, in teaching prior to inspection, a coefficient is calculated for each inspection region using an image of a non-defective substrate and registered, so at the time of inspection, for each inspection region to be processed It is possible to generate a grayscale image with a very small density difference between the color areas corresponding to the solder parts by calculation using a coefficient suitable for the area, and accurately determine whether or not the component electrode is floating. It becomes possible.

  According to the present invention, in the color image of the soldering site generated by the color highlight optical system, the variation in density among the plurality of color areas appearing in the solder portion is reduced, and the color area representing the component electrode is reduced. It can be converted into a grayscale image having a density different from that of the solder portion. Therefore, it is possible to perform a solder inspection using a color image generated by a color highlight optical system and to inspect a component electrode floating defect using the same color image as appropriate.

FIG. 1 shows a configuration of an optical system of a substrate inspection apparatus to which the present invention is applied and a main electrical configuration.
This board inspection device performs a solder inspection on each soldered part and an inspection for detecting the float of a component electrode on the printed circuit board S after the soldering process, and performs a series of processes relating to the inspection. The control processing device 1 to be executed has a configuration in which a camera 2, an illumination device 3, a substrate stage 4 and the like are connected.

  The substrate stage 4 includes a table unit 41 for supporting a substrate to be inspected, a moving mechanism 42 including an X-axis stage and a Y-axis stage (both not shown).

  The camera 2 and the illumination device 3 constitute an optical system called “color highlight system”. The camera 2 generates a color still image of the substrate to be inspected, and is arranged in a state where the imaging surface is directed downward and the optical axis is aligned in the vertical direction above the substrate stage 4.

  The illuminating device 3 includes three annular light sources 31, 32, and 33 disposed between the substrate stage 4 and the camera 2. These light sources 31, 32, and 33 emit red, green, and blue color lights, respectively, and are disposed in a state in which each central portion is aligned with the optical axis of the camera 2. The light sources 31, 32, and 33 are set to have different diameters so that light can be emitted from different directions when viewed from the substrate. Specifically, the red light source 31 has the smallest diameter and is disposed at the highest position. Hereinafter, the diameters of the green light source 32 and the blue light source 33 increase in order, and the light sources with larger diameters are arranged at lower positions.

  According to the optical system configured as described above, it is possible to generate a color image in which the state of inclination of the solder surface is reflected in the distribution state of each color region of red, green, and blue for each soldering portion on the substrate.

  The control processing apparatus 1 performs various inspections by processing the generated color image while controlling the operations of the substrate stage 4 and the camera 2 described above. The control unit 12, the illumination control unit 13, the XY stage control unit 14, the memory 15, the inspection result output unit 16, the operation unit 17, the monitor 18, and the like are connected.

  The image input unit 11 includes an interface circuit that receives three primary color image signals from the camera 2, an A / D conversion circuit that digitally converts these image signals, and the like. By this digital conversion, the color of each pixel is represented by a combination of R, G, and B gradation data (for example, a numerical range of 0 to 255 in an 8-bit configuration). The gradation data for each color is “color parameter”, and the combination of the three color parameters is “color data”. The color data of each pixel is stored in the image storage area of the memory 15 in association with the address of the pixel.

  The imaging control unit 12 controls the imaging timing of the camera 2, and the illumination control unit 13 performs lighting and extinguishing timings, light amount adjustments, and the like of the light sources 31 to 33 of the lighting device 3. The XY stage control unit 14 controls the movement timing and movement amount of the substrate stage 4.

  In addition to image data, the memory 15 stores a program used for inspection of each site to be inspected, inspection reference data, and the like. The control unit 10 executes a series of processes related to inspection using these programs and inspection reference data. Roughly described, by controlling the movement of the substrate stage 4, the camera 2 is caused to pick up an image while sequentially adjusting the field of view of the camera 2 to a plurality of positions on the substrate. Further, each time imaging is performed, an inspection region is set in the color image processed by the image input unit 11, and processing such as detection of a region to be inspected, measurement, and determination of suitability of the measurement value is executed. The result of the determination is output to an external device (not shown) or the like via the inspection result output unit 16.

  In the board inspection apparatus of this embodiment, an inspection region is set at each soldering portion in the color image, and solder inspection is sequentially performed. In this solder inspection, the color data of each pixel in the inspection area is binarized using the binarization threshold value set for each of the color parameters R, G, and B, so that the visible color is red. The three types of color regions are extracted: a region (red region), a region in which the visually recognized color is green (green region), and a region in which the visually recognized color is blue (blue region). Further, the area of these color regions is individually measured, and each measured value is collated with a determination reference value, thereby determining whether the solder surface is inclined properly.

  Furthermore, in this embodiment, for the inspection region corresponding to the mini-mold component that becomes defective when the component electrode is exposed from the solder, the color image in the inspection region is converted into a grayscale image (hereinafter simply referred to as 256 gray scale). The image is converted into a “grayscale image”), and an inspection (hereinafter, referred to as “floating defect inspection”) is performed to determine whether or not the component electrode has a floating defect using the converted image.

  In this floating defect inspection, edges are extracted from the converted grayscale image, and the number of extracted edge pixels is compared with a predetermined threshold value. When the number of edge pixels exceeding the threshold is extracted, it is considered that an edge corresponding to the component electrode appears, and it is determined that “there is a floating defect”.

The image conversion process executed in the above-described floating defect inspection is performed for each pixel in the inspection region by performing an operation using the color parameters R, G, and B constituting the color data of the pixel. A luminance value (pixel brightness) corresponding to the data is obtained, and the luminance value is converted into a value of 256 gradations and set as pixel data.
In this embodiment, the color parameters R, G, and B are applied to the solder surface where the color regions of red, green, and blue are distributed so that the color difference does not appear as a density difference in the converted image. Weighting is performed using specific coefficients α, β, and γ, and the luminance value is calculated by calculating the sum of the weighted color parameters.

  Coefficients α, β, and γ (hereinafter referred to as “brightness conversion coefficients”) used in the above calculation are non-defective models of the substrate to be inspected (hereinafter referred to as “non-defective substrate”) in the teaching mode before inspection. The color image (hereinafter referred to as “good image”) is calculated and registered in the memory 15.

  FIG. 2 shows an example of setting the areas used for calculating the luminance conversion coefficients α, β, and γ. In this example, the region 30 is set in a range in which the respective colors of red, green, and blue are distributed in the soldering portion 22 of the component 21 to be subjected to the floating defect inspection in the non-defective product image. The region 30 of this embodiment is specified by the user, but is not limited to this, and each color region of red, green, and blue is detected using a binarization threshold value for solder inspection, and the region is based on the detection result. It is also possible to set 30 automatically.

  In this embodiment, each pixel in the region 30 is classified into three groups of red, green, and blue based on visually recognized colors, and representative values of the color parameters R, G, and B (for this embodiment) Then, the average value) is calculated. Then, when the calculation for luminance conversion is executed for each group using the calculated representative value, the values of the respective luminance conversion coefficients α, β, and γ are adjusted so that the same result can be obtained by any calculation. .

Here, a method for obtaining the luminance conversion coefficients α, β, γ will be specifically described.
An average value of R, G, B values obtained for the “red” group in the region 30 is Rr, Gr, Br, and a luminance value calculated using these is Hr. Similarly, for the “green” group, the average value of R, G, B values is Rg, Gg, Bg and the luminance value is Hg, and for the “blue” group, the average value of R, G, B values is Rb, Gb. , Bb and the luminance value is Hb.

The calculation for converting pixel-unit color data into luminance values is to obtain the sum of these multiplication values by multiplying the color parameters R, G, B by coefficients α, β, γ, respectively. Therefore, equations for obtaining the luminance values Hr, Hb, Hg for each group are as shown in the following equations (1) to (3), respectively.
Hr = α · Rr + β · Gr + γ · Br (1)
Hg = α · Rg + β · Gg + γ · Bg (2)
Hb = α · Rb + β · Gb + γ · Bb (3)

  Therefore, the luminance conversion coefficients α, β are set by setting the ternary simultaneous equations according to the equations (1) to (3) on the assumption that the luminance values Hr, Hg, Hb are equal (that is, Hr = Hg = Hb). , Γ can be calculated.

According to the luminance conversion coefficient obtained by the above method, the difference in luminance value between the color regions can be made extremely small in the color image of the solder portion where each color of red, green, and blue is distributed. Therefore, the variation in the density of the solder portion can be extremely reduced even in the grayscale image after conversion.
Therefore, if the part to be inspected is in good condition, the color image in the inspection area set in the soldered part of this part is converted into a grayscale image using the coefficients α, β, and γ. In such a case, it is considered that there is no large density variation in the converted image.

  On the other hand, even if the component electrode floats poorly and the defective electrode appears in the color image with the same color (mainly red) as the solder part, the intensity of the specularly reflected light from the electrode part and the positive part from the solder part Since there is a difference between the reflected light intensities due to a difference in reflectance, the calculation results for both are different. Therefore, in the grayscale image after conversion in this case, there is a high possibility that a significant density difference occurs between the electrode portion and the solder portion.

  As described above, if the image in the inspection region set in the soldering region is converted into a grayscale image by the luminance conversion coefficients α, β, and γ obtained by the above formulas (1) to (3), the solder portion and the electrode Although there is a density difference between the portions, a grayscale image having a small density variation in the solder portion can be generated. Therefore, when the edge extraction process is performed on the grayscale image after conversion, the edge 50 of the electrode 20 appears at the portion where the component electrode 20 is exposed as shown in FIG. The edges 53 and 54 as shown in FIG. 7 do not appear at the position corresponding to the boundary between the color regions distributed in the area, and it is possible to prevent the electrode 20 from being erroneously detected.

  The luminance conversion coefficients α, β, and γ are preferably obtained for each inspection area that needs to be subjected to a floating defect inspection. If the intensity of specularly reflected light from the soldered part varies depending on the positional relationship with the lighting device, or if the tilt angle of the solder surface varies depending on the height or size of the component, the color of the solder in the color image changes. This is because it is considered that the value of the preferable luminance conversion coefficient also varies.

However, if the intensity of the specular reflection light and the variation in the tilt angle of the solder surface are negligible, the luminance conversion coefficient obtained for a specific one of the inspection target parts is set to other inspection target parts. May be registered as a common coefficient.
Further, the luminance conversion coefficient may be calculated in the inspection, not limited to the teaching mode. For example, a color image of a soldered part determined as a non-defective product is selected from the soldering parts that are not subject to the float defect inspection, and the luminance conversion coefficient α is obtained using each of the red, green, and blue colors in the color image. , Β, γ can be obtained and used for floating defect inspection.

  FIG. 4 shows a flow of processing in the teaching mode in the substrate inspection apparatus. In this example, in order to simplify the explanation, it is assumed that the substrate is imaged only once for both teaching and inspection. In the following description and FIGS. 4 and 6, each processing step is abbreviated as “ST”.

  In the teaching mode, first, a non-defective substrate is loaded onto the substrate stage 4 in the first ST101, and the non-defective substrate is imaged in the next ST102.

In ST103, the color image generated by the above imaging is displayed on the monitor 18, and the following processing of ST103 to 115 is executed for each component in the image.
First, in ST103, a range designation operation is accepted for each soldering part on the display screen for the first teaching target part, and the designated range is set as an inspection area. However, the setting of the inspection area is not limited to this, and a method of temporarily setting the inspection area using CAD data of the substrate to be processed that has been acquired in advance and accepting a correction operation by the user may be used.

  In ST104, a binarization threshold value and a determination reference value for solder inspection are set in one of the set inspection areas. The binarization threshold value allows the user to specify areas in which red, green, and blue colors appear in the inspection area, and in these areas, a representative value such as an average value in the area is set for each color parameter. It is set by the calculation method. The determination reference value is set by a method of actually extracting each color area in the inspection area using the set binarization threshold and measuring the area for each area.

  Further, when the soldering part being processed is a subject of the floating defect inspection, ST105 becomes “YES”, the process proceeds to ST106, and an operation for designating the region 30 shown in FIG. 2 is accepted.

  In the next ST107, the image in the designated area 30 is binarized using the binarization threshold value set in ST104, so that each pixel in the area is “red”, “green”, “blue”. Classify into groups.

In ST108, for each group, the average value of the color parameters R, G, B in that group is calculated.
In subsequent ST109, these average values are applied to the above-described equations (1) to (3) to calculate the luminance conversion coefficients α, β, and γ.

  Next, in ST110, a luminance value is calculated for each pixel in the inspection area using each of the luminance conversion coefficients α, β, and γ, and a color image in the inspection area is converted into a grayscale image based on the calculation result. To do. In ST111, an edge extraction filter such as a Sobel filter is applied to the converted grayscale image to calculate a density gradient for each pixel.

  In ST112, a threshold value dh for edge pixel extraction is set based on the maximum value in the density gradient calculated in ST111 (for example, dh is a value obtained by adding a numerical value corresponding to an error to the maximum value of the density gradient). To do.)

  Similarly, it is necessary to pay attention to the inspection areas set for the parts to be taught one by one, set the binarization threshold value for solder inspection and the determination reference value, and then perform the floating defect inspection. In this case, the luminance conversion coefficients α, β, γ are calculated and the threshold value dh for edge extraction is set. When the processing for all the inspection areas is completed, ST113 becomes “YES” and the process proceeds to ST114.

  In ST114, in addition to the setting data (for example, the coordinates of the upper left vertex and the lower right vertex) of each inspection area, the binarization threshold value and determination reference value for solder inspection set in ST104 are registered. Further, when performing a floating defect inspection, the brightness conversion coefficients α, β, γ calculated in ST109 and the threshold value dh for edge pixel extraction set in ST112 are registered.

  Furthermore, in the next ST115, a window for detecting a rotational deviation amount described later is set and registered. As shown in FIG. 5, this window is set as a belt-like window LW in which only an edge extending in one direction is included in a range where the electrode of the component main body 21 is not provided.

  In the same manner, various data necessary for inspection are set and registered for all parts in the image. When the processing for all parts is completed, ST116 becomes “YES” and the teaching mode is terminated.

FIG. 6 shows an inspection flow for a single substrate.
First, in ST201, a substrate to be inspected is carried into the substrate stage 4, and imaging is performed in the next ST202. As a result, a color image to be inspected is generated.

Next, in ST203, the inspection data registered in the teaching mode is read for the first inspection target part.
In the next ST204, the above-described window LW is set for this component, and the rotational deviation amount θ of the component main body is detected. Specifically, an edge pixel in the window LW is detected by an edge extraction filter or the like, a Hough transform using the coordinates of each detected pixel is executed, and a straight line that approximates the edge is specified. Further, an angle formed by the specified straight line and a predetermined reference direction is calculated, and the calculated angle is set as the rotational deviation amount θ.

Next, in ST205, the first inspection area is set. In the next ST207, the rotational deviation of the inspection region is corrected using the rotational deviation amount θ detected in ST204. Specifically, an image within a predetermined range including the inspection area is rotated so as to approach the reference direction by an angle θ while the position of the inspection area is fixed.
If the component is rotationally displaced, a portion that should originally be included in the inspection region deviates from the inspection region, or a portion that should not be included in the inspection region enters the inspection region. According to the process of ST207, such a problem is corrected and the inspection area can be set in an appropriate range.

  In the next ST207, red, green, and blue color regions are extracted by binarization processing using a binarization threshold for the corrected inspection region.

  Subsequently, in ST208, the area of each color region is measured. In ST209, the suitability of the solder surface state is determined by collating the measurement values obtained for each color region with the corresponding determination reference values.

  When inspection data for floating defect inspection is not registered in the inspection area, ST210 is “NO” and the inspection in this inspection area is terminated.

On the other hand, if inspection data for floating defect inspection is registered, ST210 becomes “YES” and the process proceeds to ST211.
In ST211, for the same inspection area where the solder inspection is executed, the color data of each pixel in this area and the luminance conversion coefficients α, β, γ registered for the area at the time of teaching are used. A luminance value is calculated. Based on this calculation result, the color image in the inspection area is converted into a grayscale image.

  In ST212, the density gradient of each pixel in the converted grayscale image is extracted. In the next ST213, the number of pixels is counted while identifying the pixels whose extracted density gradient exceeds the threshold value dh as edge pixels. In ST214, this count value is compared with an electrode detection threshold value to determine whether or not a floating defect has occurred. In this case, if the count value exceeds the threshold value, it is determined that “there is a floating defect”.

  Thereafter, in the same manner as described above, an inspection region is set for each soldering portion, a solder inspection is performed, and a floating defect inspection is further performed as necessary. When the processes for all the inspection areas are completed, ST215 is “YES”, and the process returns to ST203 after the determination process of ST216. Thereafter, the same processing is executed for each component.

  When the processing for all the parts is completed, ST216 becomes “YES”, the process proceeds to ST217, and data obtained by integrating the determination results for each part is output as an inspection result. Further, in ST218, the substrate to be inspected is unloaded and the process is terminated.

  According to the processing shown in FIG. 4 and FIG. 6, the luminance conversion coefficients α, β, and γ are calculated for each inspection region using the non-defective image in the teaching mode, and the region for each inspection region is also used in the inspection. The luminance conversion coefficients α, β, and γ registered for each are used for conversion to a grayscale image, so that the variation in the density of the portion where the color distribution corresponding to the solder portion is generated can be made extremely small. it can. Further, at the time of teaching, the inspection area in the non-defective image is converted into a grayscale image using each calculated coefficient, and the threshold value dh for edge pixel extraction is set based on the maximum value of the density gradient in the image. Even if a certain degree of density variation occurs in the grayscale image after conversion of the solder portion in the inspection region, edge pixels are not extracted due to a small density gradient caused by the variation. Therefore, the edge corresponding to the boundary between the component electrode and the solder portion can be detected with high accuracy, and the accuracy of the floating defect inspection can be ensured.

  In the above description, the exposed state of the component electrode 20 has been described as a floating defect. However, depending on the manufacturer, a part of the exposure may be permitted and determined as a non-defective product. Even in such a determination criterion, by adjusting the threshold value dh for collation of the number of extracted edge pixels, or by providing a process exclusion region in a range that allows exposure of the component electrode 20 in the inspection region, It can be easily handled.

It is a block diagram which shows the electrical structure of a board | substrate inspection apparatus with the structure of an optical system. It is explanatory drawing which shows the example of a setting of the area | region used for calculation of a luminance conversion coefficient. It is explanatory drawing which matches and shows the color image and the edge extraction result with respect to the grayscale image converted from this color image. It is a flowchart which shows the flow of the process at the time of teaching. It is explanatory drawing which shows the example of a setting of the window for a rotation deviation amount detection. It is a flowchart which shows the flow of a process of a test | inspection. It is explanatory drawing which contrasts and shows the state with the favorable mounting of minimold components, and the state where the floating defect produced. It is explanatory drawing which matches and shows the color image and the result of having converted this color image into the gray image by the conventional method, and performing the edge extraction process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Control processing apparatus 2 Camera 3 Illumination apparatus 10 Control part 15 Memory 20 Component electrode 22 Solder 50 Edge S board | substrate

Claims (4)

An illumination device that emits light of different colors from a plurality of directions with different elevation angles with respect to a substrate to be inspected, and a color image arranged at a position where regular reflected light from the substrate can be incident on illumination light from this illumination device The component mounting board after soldering is imaged by the imaging apparatus under illumination by the lighting apparatus, and the solder on the board is based on the distribution state of the colors corresponding to each illumination color in the generated image A method for inspecting the presence or absence of a floating defect in a component electrode by using a substrate inspection apparatus for inspecting an attachment site,
A function of executing a calculation for calculating the sum of each multiplication value after multiplying a plurality of color parameters constituting color data of each pixel by a predetermined coefficient, and each in the inspection area set in the color image to be processed A function for converting the color image in the inspection region into a grayscale image by replacing the color data of the pixel with the value obtained by the calculation is set in the substrate inspection apparatus,
Prior to the inspection, for the color image of the soldered part, the representative value of each color parameter in the area where the color appears for each color corresponding to each illumination color is obtained, and these representative values are used for each illumination color. Execute the process of calculating each coefficient assuming that the calculation result obtained when executing the calculation for each corresponding color is the same value,
A step of setting an inspection area for inspecting a floating defect of a component electrode at a soldering portion in a color image to be inspected;
For each pixel in the inspection area, the calculation is performed using the value of each color parameter constituting the color data of the pixel and the value of each coefficient obtained before the inspection, and based on the calculation result, the color in the inspection area Converting the image into a grayscale image;
Extracting edge pixels from the converted gray image;
When the number of edge pixels exceeding a predetermined threshold is extracted, it is determined that there is a component electrode floating defect in the inspection region.
A method for inspecting a component electrode floating defect.   The processing for calculating each coefficient before the inspection corresponds to a component to be inspected for a defective floating of a component electrode, using a color image when the imaging device images a non-defective substrate having the same configuration as the substrate to be inspected. The method for inspecting a floating defect of a component according to claim 1, which is performed for each soldering part. An illumination device that emits light of different colors from a plurality of directions with different elevation angles with respect to a substrate to be inspected, and a color image arranged at a position where regular reflected light from the substrate can be incident on illumination light from this illumination device The component mounting board after soldering is imaged by the imaging apparatus under illumination by the lighting apparatus, and the solder on the board is based on the distribution state of the colors corresponding to each illumination color in the generated image In the board inspection device that inspects the attachment part and outputs the inspection result,
A calculation means for executing a calculation for calculating a sum of each multiplication value after multiplying each of a plurality of color parameters constituting the color data of each pixel by a predetermined coefficient;
For the color image of the soldered part, obtain the representative value of each color parameter in the area where the color appears for each color corresponding to each illumination color, and use these representative values for each color corresponding to each illumination color A coefficient calculation means for executing a process of calculating each coefficient on the assumption that the calculation result obtained when the calculation is performed is the same value;
Among the inspection areas set for each soldered part in the color image generated by the imaging device, in the inspection area to be inspected for the floating defect of the component electrode, for each pixel in the area, Calculation by the calculation means using the value of the color parameter constituting the color data and the value of each coefficient calculated before the inspection by the coefficient calculation means, and the color data of the pixel by the value obtained by this calculation Image conversion means for converting a color image in the inspection region into a grayscale image by processing for replacing
Discriminating means for extracting edge pixels from the grayscale image generated by the image converting means, and determining the presence or absence of a floating defect of the component electrode based on the extraction result;
Output means for outputting the discrimination result by the discrimination means,
The discrimination means compares the number of extracted edge pixels with a predetermined threshold value, and when an edge pixel exceeding the threshold value is extracted, the component electrode floats in the inspection region corresponding to the grayscale image. A substrate inspection device that determines that there is a defect. The coefficient calculation means performs a process for obtaining each coefficient using a color image obtained when the imaging apparatus images a non-defective substrate having the same configuration as the substrate to be inspected for each inspection area to be processed by the image conversion means. Run,
The substrate inspection apparatus according to claim 3, further comprising a registration unit that stores the coefficient calculated by the coefficient calculation unit for each inspection region to be processed by the image conversion unit.

Images