JP2006322951A - Method and device for surface condition inspection, and device for inspecting substrate - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable execution of reliable inspection even if brightness and tone of an image varies when the state of an inclined plane is inspected. <P>SOLUTION: An image-pickup is performed from above under irradiation with each chromatic light of red, green, and blue from respectively different elevation angles; when inspecting a solder fillet 53 by using the acquired image, 3-D data by each gradation of the R, G, B is converted to one-imensional hue data; and then a direction that the hue changes along the order of the red, green, and blue is extracted using the hue data. The extracted direction points out a direction that the gradient of the fillet changes, which is compared with a preregistered reference direction and therefore the proprietary of the inclined plane is determined. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a method for inspecting the quality of a surface state using an image obtained by imaging an object to be inspected using an object whose surface has a high specularity and includes an inclined surface on the surface, and Relates to the device. The present invention also relates to an apparatus for inspecting the surface state of the solder at each soldering site on the board, with a board on which a plurality of soldering sites are set as an inspection target.

  The applicant has previously developed an apparatus for automatically inspecting a soldering site on a substrate by an image processing technique using the specular reflectivity of the soldering site (see Patent Document 1).

Japanese Patent Publication No.6-1173

  FIG. 16 shows the configuration of the substrate inspection apparatus disclosed in Patent Document 1 and the principle of inspection. This inspection apparatus generates an image to be inspected by the three light sources 8, 9, and 10 having red, green, and blue color lights and the imaging apparatus 3. In the figure, 1 is a substrate, 2 is a solder to be inspected, and the light sources 8, 9, and 10 are arranged in different elevation angles with respect to the substrate 1 and the solder 2. On the other hand, the imaging device 3 is arranged so as to image the solder 2 from a position directly above.

  According to the above configuration, the colored light from each of the light sources 8, 9, 10 is applied to the surface of the solder 2 from different directions. Further, when light from a direction symmetric to the direction of the image pickup device 3 is irradiated with respect to the inclined surface of the solder with the direction orthogonal to the inclination as an axis, the specular reflected light of the light is guided to the image pickup device 3. That is, the color of the specular reflected light incident on the image pickup device 3 from the solder surface varies depending on the inclination of the solder surface.

  In the optical system shown in FIG. 16, the elevation angle of the red light source 8 is the largest and the elevation angle of the blue light source 10 is the smallest when viewed from the solder 2, and the green light source 9 is positioned between these light sources 8 and 10. As shown, each light source 8, 9, 10 is arranged. Therefore, if the solder 2 is a spherical solder, as shown in FIG. 17, the flat surface at the center is a red image region, the steeply inclined surface near the substrate surface is a blue image region, and these A relatively gentle inclined surface (warm inclined surface) located in the middle of each of the images is generated as a green image region, and an image cut out is generated.

  When inspecting a solder fillet, as shown in FIG. 18, the upper steeply inclined surface is blue, the intermediate warm inclined surface is green, the substrate surface near the flat surface is red, and so on. The image area by each color is cut out. Note that S1 in FIG. 18 is a land formation range.

  In this manner, in the solder image, an image divided into red, green, and blue color regions is generated in accordance with the change in the gradient of the surface. Therefore, each color pattern in the image of the solder with a good shape is registered in advance, and the quality of the solder surface state is determined by comparing each color pattern on the image to be inspected with the registered pattern. be able to.

  In the conventional inspection apparatus, based on the above principle, a predetermined binarization threshold value is set in advance for each of the R, G, and B gradation data constituting the digital color image, and obtained by the imaging apparatus 3. By binarizing the image with these binarization threshold values, red, green, and blue color regions are extracted. In addition, for these color areas, the position, shape, size, etc. of the area extracted in a good-shaped solder image are registered in advance, and the color area extraction result in the image to be inspected is registered. In this way, the quality of the solder surface state is determined.

  By the way, in the optical system shown in FIG. 16, when the brightness and hue of the illumination light vary depending on the temperature and usage time, it is considered that the brightness and hue on the image also vary. Such variations in image brightness and hue may affect the extraction result of the color region. In particular, in the solder inspection, even if the surface state is good, if the feature amount of each color region varies due to a change in hue, the reference value for a non-defective product cannot be cleared, and it may be determined as defective.

  Therefore, it is desirable to frequently change the binarization threshold in order to maintain inspection accuracy. However, in this type of inspection apparatus, it is required to repeat the inspection by receiving the substrate one after another. If the inspection is interrupted by updating the binarization threshold value, the processing efficiency may be lowered. It is also difficult to accurately determine the timing for updating the binarization threshold.

  On the other hand, as a process replacing the above binarization process, there is a method of executing a pattern matching process using an image of a board with a good mounting state as a model image. This method has an advantage that the normalized correlation calculation is performed between the image to be inspected and the model image, so that even if the illumination fluctuates, it can be handled without changing the threshold value for discrimination. However, there is a problem that even if the shape of the solder to be inspected is slightly changed, the correlation value is lowered and the correct answer rate for the discrimination processing is poor.

  Further, in recent boards, there is a tendency that parts are miniaturized and a large number of parts are mounted. Therefore, it is necessary to adjust the land area for each part according to the peripheral mounting density. For this reason, even if it is the same kind of component, the size of the land differs between the case where a relatively small number of components are mounted around and the case where a large number of components are mounted around. In response to this, the gradient of the solder fillet also varies depending on the mounting position of the component, so the size and shape of each color region in the fillet image also differ.

  For example, if there are relatively few parts mounted around, the land will be formed larger for soldering stability, so the fillet gradient will be relatively gentle, and red, green, and blue colors. The area appears moderately. On the other hand, in parts mounted in areas with high mounting density, the land area must be reduced, so the slope of the fillet becomes steep, the blue area predominates on the image, and the green and red areas It becomes difficult to appear. In this way, even in the same type of component, the distribution state of the color regions appearing on the image is different, so it is difficult to ensure inspection accuracy unless the determination reference value for each color region is set for each component. It becomes.

  The present invention has been made paying attention to the above-mentioned problem. When inspecting the state of an inclined surface, it is possible to perform a highly reliable inspection even if the brightness and hue of an image fluctuate. The first purpose.

  Furthermore, the present invention makes it possible to perform a highly accurate inspection even on an inspection object in which the gradient is steep or, conversely, the color region corresponding to all the light sources hardly appears. This is the second purpose.

  In the surface condition inspection method according to the present invention, a plurality of light sources that emit different colored lights are arranged and illuminated with respect to an inspection object having an inclined surface on the surface so that elevation angles viewed from the inspection object are different. The surface state of the inspection object is inspected using an image obtained by an imaging means installed above the arrangement position. In the present invention, a first step of setting an inspection region including an inspection object in an image and extracting a direction in which a hue change corresponding to the arrangement of each light source occurs in the inspection region; A second step of comparing the direction extracted in the step with a reference direction set in advance, and a third step of determining the quality of the surface state of the inspection object based on the comparison result of the second step And execute.

  When a plurality of light sources that emit different colored lights are arranged and illuminated so that the elevation angles viewed from the inspection object are different, the magnitude of the gradient is obtained at the portion corresponding to the inclined surface in the image obtained by the imaging means. Depending on the color appears. Further, when the slope of the inclined surface gradually changes, such as the flat surface, warm inclined surface, and steeply inclined surface in the examples of FIGS. 17 and 18, the color corresponding to each light source in the direction of the change. Can be considered to appear side by side in the order of the arrangement of the light sources as viewed from the inspection object (in order of the magnitude of the elevation angle).

  In the present invention, the “change in hue corresponding to the arrangement of each light source” means the color (emission color) emitted from each light source in the order of the arrangement of each light source as viewed from the inspection object (the one with the smallest elevation angle to the larger one). It can be said that the hue changes along the arrangement direction when arranged along the line. For example, according to the optical system shown in FIG. 16, it can be understood as a change in hue (hue) when colors are arranged in the order of red, green, and blue, or in the order of blue, green, and red.

  As described above, in the image of the portion corresponding to the inclined surface, the color corresponding to each light source is arranged in the order of the arrangement of each light source viewed from the inspection object along the direction in which the magnitude of the gradient changes. Appears at That is, the “change in hue corresponding to the arrangement of each light source” appears in this arrangement direction. Therefore, the direction in which the hue change corresponding to the arrangement of each light source occurs is the direction in which the gradient of the inclined surface changes (the direction from a gentle gradient to a steep portion, or a sharp gradient). It can be considered to indicate a direction from a gentle part to a gentle part.

  It can be considered that the direction in which the hue changes corresponding to the arrangement of the light sources does not change even if the brightness or hue on the image varies. Therefore, using a model object image in advance, set a reference direction that causes a hue change corresponding to each light source array, and compare the direction extracted from the image to be inspected with the reference direction. By doing so, it is possible to determine whether or not the direction in which the gradient of the inclined surface changes is appropriate.

  Further, even if the inspection object is of the same type, the gradient direction is considered not to change even if the gradient size varies. Therefore, if the reference direction is set for one of these inspection objects, the setting data can be applied to other inspection objects, and the inspection information can be efficiently taught. .

  In a preferred aspect of the above surface condition inspection method, the plurality of light sources are arranged so that the respective emission colors are aligned along the direction of hue change on the hue circle with respect to the inspection object. For example, when three light sources that respectively emit red light, green light, and blue light are used as the light source, each of the elevation angles increases in the order of red, green, blue, blue, green, and red. A light source will be arranged. Further, in the first step, for each pixel in the inspection area, one-dimensional hue data whose value changes along the hue circle is acquired, and the hue data changes to a large value or a small value. The changing direction is extracted.

  In the above aspect, since the plurality of light sources are arranged in the order along the direction of the hue change on the hue ring, the change in hue corresponding to the arrangement of each light source on the image also corresponds to the change in the hue ring. Become. Further, by using one-dimensional hue data whose value changes along the hue circle, the direction in which the hue change corresponding to the arrangement of each light source occurs is changed to a direction in which the hue data changes to a large value or a small value. It can be expressed as a direction that changes to a value. Therefore, according to this aspect, the direction in which the hue change occurs can be easily extracted as a vector based on the hue change amount.

  In another preferable aspect of the surface state inspection method, in the first step, for each pixel in the inspection region, the degree of intensity of the color component corresponding to any one of the plurality of light sources is reflected. The obtained one-dimensional hue data is extracted, and the direction in which the hue data changes to a large value or the direction to change to a small value is extracted.

  For example, when the optical system shown in FIG. 16 is used, paying attention to the blue component corresponding to the blue light source 10, that is, the gradation data B, the gradation data B becomes larger as the gradient becomes steeper, and the gradient becomes larger. It can be considered that it becomes smaller as it becomes gentler. Therefore, for the gradation data B, a ratio (for example, B / R or B / (R + G)) or a ratio (B / (R + G + B)) to the lightness is obtained with respect to the other color components (gradation data R, G). This can be used as hue data. Also in this case, the direction in which the hue change corresponding to the arrangement of each light source occurs can be expressed as the direction in which the hue data changes to a large value or the direction in which the hue data changes to a small value. The generated direction can be easily extracted as a vector based on the amount of change in hue.

  When acquiring one-dimensional hue data and extracting the change as in the above two aspects, in the first step, paying attention to the plurality of pixels in the inspection region in order. Extracting a hue data change amount for each of a plurality of directions passing through the pixel of interest; averaging a hue data change amount extracted for each pixel of interest for each direction to set a plurality of vectors; and And a step of setting a composite vector of the vectors. According to this aspect, in the inspection region, for each of a plurality of directions, a vector whose length is the average amount of change in hue in each direction is obtained, and then these vectors are combined to obtain the hue. It is possible to extract a hue change along the ring, that is, a direction in which a hue change corresponding to the arrangement of each light source occurs.

  It should be noted that the combined vector extracted in the above-described manner preferably has a length equal to or longer than a predetermined threshold according to the purpose of inspection. When only a composite vector shorter than the threshold value is extracted, it is desirable to consider that there is no good inclined surface in the inspection region.

  Next, the surface condition inspection apparatus according to the present invention includes a lighting unit in which a plurality of light sources that emit different colored lights are arranged so as to have different elevation angles when viewed from the arrangement position of the inspection object, and the inspection object An image input means arranged above the arrangement position, an image input means for capturing an image generated by the image pickup means with each light source of the illumination means turned on, and an image input by the image input means When an inspection area including the inspection object is set, and in the inspection area, the extraction means for extracting the direction in which the hue change corresponding to the arrangement of each light source occurs, and the surface state of the inspection object is good A registration unit for registering a reference direction extracted by the extraction unit, and a direction of change extracted by the extraction unit is compared with a reference direction registered by the registration unit. Comprising a determining means for determining the quality of the surface state of the object, and output means for outputting a discrimination result by said discrimination means.

  In the above, the illumination means may be, for example, a ring-shaped light source having a different diameter for each color and arranged centering on the center. Further, a plurality of light emitters (such as LEDs) that emit light of the same color arranged in a ring shape can be used as one light source. The light source is not limited to a ring shape, and a line light source can also be used.

  The imaging means can be constituted by a CCD camera capable of generating an image signal for each color. When a ring-shaped light source is used as the illumination unit, the imaging unit is preferably arranged in the vertical direction with the optical axis aligned with the central axis of each light source.

  The image input unit, the extraction unit, the registration unit, the determination unit, and the output unit can be provided in the same casing. Among these, the extraction means, the registration means, and the determination means can be configured by a computer in which a program for executing the processing of these means is incorporated. These computers can be provided for each means, and each means can be constituted by one computer.

  The image input means can include an amplifier circuit for amplifying the image signal from the imaging means, an A / D conversion circuit for digitally converting the image signal, and the like. However, when a digital camera is used as the imaging means, the image input means can be configured as an input port for capturing digital color image data.

  The extracting means is for executing the first step described above, and the determining means is for executing the second and third steps. The registration unit is configured to register a reference direction extracted from the image in a computer memory when an image of a non-defective model of the inspection object generated using the illumination unit or the imaging unit is input. can do. For example, when an operation for setting an inspection area on an image of a non-defective model is performed, image data in the area is processed by the extraction unit, and the extracted direction is stored in a memory as a reference direction. , A registration means. Alternatively, a means for accepting an operation for designating a direction in which colors corresponding to each light source are arranged in an order corresponding to the arrangement of the light sources on the non-defective model image, and storing the direction in the memory as a reference direction is used as a registration means. Also good.

  The output means can be configured as a display interface for displaying the determination result by the determination means on a display device, or an output interface for outputting the determination result to an external device. In addition, the output means can be configured as means for recording the determination result on a storage medium such as a flexible disk, a CD-R, or a magneto-optical disk.

In the surface condition inspection apparatus according to a preferred aspect, the light sources of the illuminating means are arranged so that the respective emission colors are aligned along the direction of hue change on the hue circle with respect to the inspection object. In addition, the extraction unit acquires one-dimensional hue data whose value changes along the hue circle for each pixel in the inspection region, and the hue data changes to a large value or a small value. It is set to extract the direction to perform.
In a preferred configuration of the illumination means, three light sources that respectively emit red light, green light, and blue light are arranged in such an order that the elevation angle increases in the order of red, green, blue or blue, green, and red. The

  In the surface condition inspection apparatus according to another preferred aspect, the extracting means is one-dimensional hue data reflecting a degree of intensity of a color component corresponding to one of the plurality of light sources for each pixel in the inspection region. And the direction in which the hue data changes to a large value or the direction to change to a small value is extracted. Also in this aspect, the illuminating means includes three light sources that emit red light, green light, and blue light in the order of increasing the elevation angle in the order of red, green, blue or blue, green, red. It is desirable to arrange. When the illumination means has such a configuration, R gradation data corresponding to a red light source or B gradation data corresponding to a blue light source is selected as a color component used for the hue data. desirable. In this case, the selected gradation data can be used as the hue data by obtaining a ratio with respect to the other gradation data, or a ratio with respect to the sum or brightness of the other gradation data. For example, if the selected gradation data is B, B / R or B / (R + G)) or B / (R + G + B) can be obtained as the hue data.

  Further, the extracting means focuses on the plurality of pixels in the inspection region in order, and extracts the amount of change in hue data for each of a plurality of directions passing through the pixel of interest, and the hue extracted for each pixel of interest. Means for averaging a change amount of data for each direction and setting a plurality of vectors, and means for setting a combined vector of these vectors can be included.

  Next, the present invention can be applied to a board inspection apparatus for inspecting a board on which a plurality of components are soldered. This board inspection apparatus comprises the same illumination means, imaging means, and image input means as those of the surface condition inspection apparatus described above, and sets inspection areas at the soldering sites in the image input by the image input means. Extraction means for extracting the direction in which the hue change corresponding to the arrangement of each light source occurs for each inspection area, and the extraction means when the surface condition of the solder in the inspection area is good A registration means for registering a reference direction and a process for comparing the direction of change extracted by the extraction means with the reference direction registered by the registration means for each inspection region, Discriminating means for discriminating between and an output means for outputting a discrimination result by the discriminating means.

  The extraction means, registration means, discrimination means, and output means of the substrate inspection apparatus can basically have the same configuration as that of the surface condition inspection apparatus. Moreover, the same aspect as the said surface state inspection apparatus can be set also to this board | substrate inspection apparatus.

Furthermore, in the board inspection apparatus according to a preferred aspect, when a plurality of components of the same type are mounted on the substrate to be inspected, the registration unit registers with respect to one soldering portion of these components. It has the function of applying the reference direction to other parts of the same type. According to this function, even when a plurality of components of the same type are mounted on the board, it is not necessary to perform registration processing for each of these components, and the efficiency of teaching inspection information can be improved.
The reference direction of the registration target may be set by accepting a designation operation by the user in the same manner as described for the surface condition inspection apparatus, or by causing the extraction means to process an image of a non-defective substrate. The reference direction may be extracted. Also, when the reference direction is diverted to another part of the same type, if the mounting direction of the part is different from the previously registered part, the reference direction is corrected using the angular difference of the mounting direction. Processing is required.

  According to the board inspection apparatus according to the present invention, it is possible to accurately determine the suitability of the solder inclination state by determining whether or not the direction of change in the gradient of the inclined surface of the soldering portion is the correct direction. . It is also possible to determine whether or not the solder inclined surface is formed and whether or not the formation position is appropriate based on the magnitude of the hue change corresponding to the arrangement of each light source and the extracted position of the hue change. In addition, when multiple identical components are mounted on the same board, the reference direction does not change even if the solder gradient changes depending on the density of the components at the mounting position. It becomes possible to execute.

  In this invention, by extracting the direction in which the hue change corresponding to the arrangement of each light source occurs on the image to be inspected, the direction in which the gradient of the inclined surface changes is obtained, and the direction is registered as a reference standard. Since it is compared with the direction, stable inspection can be performed even if the brightness and color tone on the image fluctuate. In addition, even if the brightness and color tone on the image fluctuate, there is no need to change the registered reference direction. This is convenient when a plurality of inspection objects must be processed one after another, such as a board inspection. Can increase the sex. In addition, by extracting the direction in which the slope of the inclined surface changes, for an inspection object in which the color area corresponding to all the light sources is unlikely to appear due to a steep or gentle slope, A highly accurate inspection can be performed.

FIG. 1 shows the configuration of a substrate inspection apparatus according to an embodiment of the present invention.
This board inspection apparatus is for processing an image obtained by imaging a board to be inspected to determine whether or not a soldering site on the board or a component mounting state is appropriate. The optical unit 4, the control processing unit 5, the X-axis table unit 6, the Y-axis table unit 7, and the like are included.
In addition, 1T in the figure is a substrate to be inspected (hereinafter referred to as “inspected substrate 1T”). Further, 1S is a reference board with a good soldering state and component mounting state, and is used for teaching prior to inspection.

  The Y-axis table unit 7 includes a conveyor 7A that supports the substrates 1S and 1T. The conveyor 7A is moved by a motor (not shown) to move the substrates 1S and 1T in the Y-axis direction (a direction orthogonal to the drawing sheet). ). The X-axis table unit 6 supports the imaging unit 3 and the light projecting unit 4 above the Y-axis table unit 7 and moves them in the X-axis direction (left-right direction in the figure).

  The light projecting unit 4 is composed of three annular light sources 8, 9, and 10 having different diameters. These light sources 8, 9, and 10 emit red, green, and blue color lights, respectively, and are positioned at a position directly above the observation position to thereby have different elevation angles as viewed from the substrates 1S and 1T. It is arranged so that it may be located in the direction corresponding to.

  The image pickup unit 3 is a CCD camera for generating a color image, and is positioned so that its optical axis corresponds to the center of each light source 8, 9, 10 and is along the vertical direction. As a result, the reflected light from the substrates 1S and 1T to be observed enters the imaging unit 3, is converted into color signals R, G, and B of the three primary colors and is input to the control processing unit 5.

  The control processing unit 5 includes a computer including a CPU as a control unit 11, and an image input unit 12, a memory 13, an imaging controller 14, an image processing unit 15, an illumination control unit 16, an XY table controller 17, an inspection unit 18, and a teaching table 19. , A library data storage unit 20, a data management unit 21, an input unit 22, a CRT display unit 23, a printer 24, a transmission / reception unit 25, an external memory device 26, and the like.

  The image input unit 12 includes an amplification circuit that amplifies the three primary color image signals from the imaging unit 3, an A / D conversion circuit that converts these image signals into digital signals, and the like. In the memory 13, in addition to digital grayscale image data for each color (hereinafter referred to as "gradation data R, G, B"), a binary image obtained by processing these gradation data R, G, B is stored. Data, hue data, etc. are stored.

  The imaging controller 14 includes an interface for connecting the imaging unit 3 to the control unit 11. The imaging controller 14 drives the imaging unit 3 based on a command from the control unit 11 and adjusts the output level of each color light. Control. The illumination control unit 16 is for adjusting the light quantity of each light source of the light projecting unit 4. In this embodiment, the light amounts of the light sources 8, 9, and 10 are adjusted so that white illumination is performed by mixing red, green, and blue color lights.

  The XY table controller 17 includes an interface for connecting the X-axis table unit 6 and the Y-axis table unit 7 to the control unit 11, and based on commands from the control unit 11, the X-axis table unit 6 and the Y-axis table unit 7. Controls the movement movement.

  The teaching table 19 is a storage unit for storing board inspection information. In this inspection information, in addition to the setting conditions of inspection windows to be described later, binarization threshold values (R, G, B) for extracting the colors of the inspection objects such as parts, lands, and solder, respectively. In addition to the binarization threshold value for each, a binarization threshold value for lightness is included.) And a determination reference value (position for determining the quality of the pattern feature extracted by these binarization threshold values) , Etc., are set for each type of feature quantity such as size). Furthermore, in this embodiment, reference data for a hue change vector, which will be described later, is registered as inspection information for solder fillets.

  The library data storage unit 20 is a storage unit for registering reference inspection information (hereinafter referred to as “library data”) for each component type. In this embodiment, setting data for an inspection window, which will be described later, the type of inspection program executed in each inspection window, and the like are registered as library data.

  Various types of inspection information registered in the teaching table 19 are taught using an image obtained by imaging the reference substrate 1S and the library data prior to inspection, and are determined for each type of substrate. Combined as a file. The data management unit 21 is a memory in which link information for associating a board type with a determination file is stored. After receiving the input of the board name of the board 1T to be inspected, the control unit 11 reads the determination file corresponding to the board 1T to be inspected based on the link information of the data management unit 21 and sets it in the memory 13. The image processing unit 15 and the inspection unit 18 execute processing based on the inspection information in the read determination file.

  The image processing unit 15 processes the gradation data R, G, and B stored in the memory 13 at the time of inspection, and extracts various parameters for determining pass / fail. The inspection unit 18 compares each extracted parameter with the corresponding inspection information, and determines, for each mounted component, whether a component is missing or misaligned, whether a soldered part is good or bad, and the like. The control unit 11 determines whether the inspected substrate 1T is a non-defective product by combining these determination results. The final determination result is output to the CRT display unit 23, the printer 24, or the transmission / reception unit 25.

  The input unit 22 is used to input various conditions for inspection, input of inspection information, and the like, and includes a keyboard and a mouse. The CRT display unit 23 (hereinafter simply referred to as “display unit 23”) receives supply of image data, inspection results, and the like from the control unit 11 and displays them on the display screen. The printer 24 receives the inspection result from the control unit 11 and prints it out in a predetermined format.

  The transmission / reception unit 25 is for exchanging data with other devices. For example, for the substrate 1T to be inspected determined to be defective, the transmission / reception unit 25 transmits the identification information and the content of the defect to the subsequent correction device. Thus, the defective portion can be promptly corrected. The external memory device 26 is a device for reading / writing data from / to a storage medium such as a flexible disk, a CD-R, or a magneto-optical disk. The external memory device 26 stores the inspection results and stores programs and setting data necessary for the inspection. Used to capture from.

  In the above configuration, the image processing unit 15 and the inspection unit 18 are configured by a dedicated processor in which a program for executing the above-described processes is incorporated. However, a dedicated processor is not necessarily provided, and the functions of the image processing unit 15 and the inspection unit 18 may be provided to the control unit 11.

Further, in this embodiment, the light projecting unit 4 is constituted by the three annular light sources 8, 9, and 10, but instead of the light projecting unit 4, a light projecting unit 4A having a configuration as shown in FIG. May be used.
The light projecting portion 4A is divided into three housing spaces 43R and 43G arranged in the radial direction by dividing the inside of a housing 41 having a cylindrical opening 42 in the center portion by using two wall portions 45 and 46. , 43B are formed. A red LED 44R is arranged in the innermost accommodation space 43R, a green LED 44G is arranged in the central accommodation space 43G, and a blue LED 44B is arranged in a plurality of rows in the outermost accommodation space 43B. The lower surface of the housing 41 is constituted by a dome-shaped diffusion plate 47, and the imaging unit 3 is disposed above the opening 42 with the optical axis aligned in the vertical direction.

  The red light, the green light, and the blue light generated in the respective accommodation spaces 43R, 43G, and 43B are transmitted to the substrate 1T (1S) after passing through the diffusion plate 47. Accordingly, similarly to the light projecting unit 4 in FIG. 1, each color light can be irradiated to the substrate 1T (1S) from different elevation angles.

  The details of the inspection executed by the substrate inspection apparatus will be described below. In the following description, the inspection target is limited to chip components that are frequently mounted on a substrate. However, it is needless to say that components other than chip components can be inspected in the same manner.

FIG. 3 shows an example of setting an inspection window for a chip component. In the figure, W11 and W12 are windows for solder inspection, and are set so as to include lands on the board. Hereinafter, the windows W11 and W12 are referred to as “land windows W11 and W12”. Further, W2 is a window for inspecting a missing part or a positional deviation of the component, and this is hereinafter referred to as a “component window W2”.
Further, a third inspection window W3 including the land windows W11 and W12 and the component window W2 is set. The window W3 serves as a reference for positioning the land windows W11 and W12 and the component window W2. Hereinafter, this window is referred to as a “reference window W3”.

  Information indicating the setting conditions of the inspection windows W11, W12, W2, and W3 (window size, setting position, and the like) is obtained by processing of the control unit 11 that links the board design data (CAD data) and the library data. Created automatically. The CAD data includes the mounting position of each component and the identification information (component type name, model number, etc.) of the mounted component, and the library data also associates identification information such as the component name with various inspection information. Take the configuration. The control unit 11 applies the library data of the component mounted based on the identification information to the mounting position of each component indicated by the CAD data, and the specific size of each inspection window W11, W12, W2, W3. Determine the position and setting position. The sizes and setting positions of the inspection windows W11, W12, and W2 determined here are registered in the teaching table 19 as the setting conditions.

  At the time of inspection, after setting each inspection window W11, W12, W2, W3 based on the registered setting condition, the land is extracted by the reference window W3. Based on the extraction result, the setting positions of the land windows W11 and W12 and the component window W2 are finely adjusted.

  The land can be extracted by performing a binarization process for extracting the color of the land and then projecting the binary image in the x and y axis directions. Alternatively, the image in the reference window W3 may be converted into a monochrome image to extract an edge, and the edge corresponding to the land may be extracted from the extraction result.

  When each inspection window W11, W12, W2 is set, an inspection based on the inspection information registered for each of these windows is executed. For the component window W2, as in the prior art, a color pattern of a component is extracted using the binarized threshold value, and an inspection using a feature quantity such as the area and the center of gravity of the pattern is executed. Therefore, in the following, a detailed description of the fillet inspection executed in the land windows W11 and W12 will be developed.

  When inspecting the fillet, as shown in FIG. 4, the range from the edge of the component electrode 52a to the outer edge of the land window W11, 12 in the images in the land windows W11, W12 (in the figure, (A range with halftone dots) is the processing target. Hereinafter, the areas 31 and 32 corresponding to the range of the halftone dots are referred to as “inspection areas 31 and 32”. In FIG. 4, the image of the component body of the chip component is 51a and the image of the electrode is 52a, following FIG.

  FIG. 5 shows the distribution state of the color appearing on the image (upper stage) and the state when the actual shape is viewed from the side (lower stage) with respect to the solder fillet of the chip part. In FIG. 5, the component main body is 51, the electrode is 52, the fillet is 53, and these images are shown as 51a, 52a, and 53a.

FIG. 5A shows an example where the shape of the fillet 53 is good. In the fillet 53 in this case, a portion close to the component 51 is a steep slope, a portion close to the substrate 1 is a flat surface, and an intermediate portion between them is a warm inclined surface. along the direction P _a, to change red, green, in the order of blue. On the other hand, FIG. 5B shows an example of a defective case, and due to poor wettability and excessive solder, the fillet 53 is not formed, and a defective portion 54 in which the solder swells up is formed. In such a defective portion 54, a portion close to the component 51 is a flat surface, a portion close to the substrate 1 is a steep slope, and an intermediate portion therebetween is a warm slope. Accordingly, the color of the image 54a of the defective portion changes in the order of red, green, and blue along the direction P _B that is opposite to that of the non-defective product, that is, the direction P _B from the upper end to the lower end.

  As described above, the distribution state of red, green, and blue colors of the soldered portion on the image varies depending on the inclination state of the solder. However, in any case, the color changes in the order of red, green, and blue from the smaller gradient of the inclined surface to the larger gradient. This change is the same as the change along the hue circle.

Accordingly, for each pixel in the inspection regions 31 and 32, one-dimensional hue data whose value decreases so that the redness becomes stronger and the blueness becomes stronger is obtained, and the change amount of the hue data in each direction is obtained. In the direction in which the color changes in the order of red, green, and blue as in the directions P _A and P _B in FIG. On the other hand, the hue change amount is smaller than the predetermined value in the direction where the color change does not appear. In the direction in which the color changes in the order of blue, green, and red, it can be considered that the amount of change in hue is a negative value.

  Therefore, in this embodiment, a vector indicating the direction in which the hue change from the red side to the blue side appears is defined as a hue change vector, and the hue change vector is extracted using the one-dimensional hue data, Whether the fillet is good or bad is determined using the magnitude and direction of the vector.

In order to extract the hue change vector, the image processing unit 15 converts three-dimensional image data composed of R, G, and B gradation data into one-dimensional hue for each pixel in the inspection regions 31 and 32, respectively. Convert to data H. In this conversion process, the values of R, G, and B are compared with each other, and depending on whether the maximum value V _max among them is R, G, or B, the following (1) (2) (3) Select one of the expressions to perform the operation. In addition, rr, bb, and gg in the equations (1) to (3) are obtained by equations (4), (5), and (6), respectively. In (4) to (6), V _min is the minimum value among R, G, and B.

When R = V _max H = (bb−gg) * π / 3 (1)
When G = V _max H = 2 + (rr−bb) * π / 3 (2)
When B = V _max H = 4 + (gg−rr) * π / 3 (3)

_{rr = (V max -R) /} (V max -V min) ··· (4)
_{gg = (V max -G) /} (V max -V min) ··· (5)
_{bb = (V max -B) /} (V max -V min) ··· (6)

Assuming that R, G, and B are represented by 256 gradations (8 bits), the value of the hue data H obtained by the above calculation theoretically changes in the range of 0 to (4 + π / 3) (R When H = 255 and G = B = 0, H becomes the minimum value 0, and when R = G = 0 and B = 255, H becomes the maximum value 4 + π / 3.
Further, in this embodiment, the value of H is corrected using the following equation (7) so that the range of the numerical change is 0-10.
H1 = H * 10 / (4 + π / 3) (7)

  The above hue data H and H1 become smaller as the redness becomes stronger, and becomes larger as the blueness becomes stronger. Therefore, in the inspection areas 31 and 32, when a change exceeding a predetermined threshold appears in the hue data H1, the direction of the change can be set as the direction of the hue change vector.

  In this embodiment, based on the above principle, a hue change vector is extracted by a procedure as shown in FIG. Hereinafter, the contents of the processing in FIG. 6 will be described in detail with reference to the specific examples shown in FIGS. In addition, in FIG. 6, each step is shown with the code | symbol of ST101-107. In the following description, “step” is also abbreviated as “ST”. 7 to 13, the direction from left to right is the positive direction of the x-axis, the direction from bottom to top is the positive direction of the y-axis, and the horizontal direction of the component is along the x-axis direction. .

  FIG. 7 shows an example of calculation of the hue data H1 in the inspection region 31 set in the right land window W11 for the chip part 51. In order to simplify the explanation, in this example, the size of the inspection region 31 is 5 pixels in the x-axis direction and 7 pixels in the y-axis direction.

In the process of extracting the hue change vector in FIG. 6, first, in ST101, a 3 × 3 pixel mask M as shown in FIG. 8 is initially set at the upper left corner of the inspection area. Hereinafter, the nine pixels g in the mask M are represented by coordinates (i, j) (i, j = 1, 2, 3) based on the row / column relationship. Each pixel g is also referred to as a pixel g _ij using the respective coordinates.

In the next ST102, for each pixel g _ij in the mask M, a difference (g _ij −g ₂₂ ) in hue data from the center pixel g ₂₂ is calculated. Of these, values other than the operation result between the pixel g ₂₂ becomes to represent the color variation in the surrounding eight directions as viewed from the pixel g _22. The calculated value obtained for each pixel g _ij is stored in the memory 13 in association with the coordinates (i, j) of the pixel in the mask M.

  Subsequently, the process of calculating the hue change amount at each position is performed while moving the mask M pixel by pixel until the mask M is aligned with the lower right of the inspection region 31 (ST102 to ST104).

  FIG. 9 shows a configuration of hue data extracted from the mask M at each position when the mask M is scanned on the inspection area 31 having the data configuration shown in FIG. In the following, assuming that the mask M is scanned along the y-axis direction, scanning performed by aligning the mask M with the left edge of the inspection region 31 is referred to as “first-row scanning”. Scanning performed by shifting pixels to the right is referred to as “second column scanning”, and scanning performed by shifting the mask M further to the right by one pixel is referred to as “third column scanning”.

  FIG. 10 shows the result of performing the process of ST102 on the hue data extracted by the mask M in the scanning of the first column. FIG. 11 shows the same processing results as FIG. 10 for the second row scanning, and FIG. 12 shows the same processing results for FIG.

  In this way, by executing ST102 at each position while moving the mask M pixel by pixel, 15 sets of data arrays representing hue change amounts in the eight directions are extracted. When the process at the final position is completed, ST103 is “YES”, the process proceeds to ST105, and a process for obtaining an average value for each direction of the hue change amount is executed.

In ST105, specifically, the average value I _ij is calculated by grouping data included in the 15 sets of data arrays for each pixel having the same relative coordinate in the mask M. For example, regarding the pixel g ₁₁ at the upper left corner of the mask, the amount of change obtained for the pixel g ₁₁ at each scanning position (−1, +1, +1, +2, +2 in the first column, −1, + 1, + 2, + 4, + 5,3 row + 1, + 2, + 1, + 3, by obtaining the average value _{I 11} of + 4), it is possible to obtain a value of + 1.8. Similarly, average values I _{12 to} I ₃₃ are obtained for the other pixels g _{12 to} g ₃₃ .

FIG. 13 shows the calculation result of the average value I _ij obtained by the process of ST105 from the 15 sets of data arrays. According to this result, the average values I ₁₁ , I ₂₁ , I _{31 in} the pixels g ₁₁ , g ₂₁ , g ₃₁ on the left side of the central pixel g ₂₂ are positive, and the pixel g on the right side of the pixel g ₂₂ is positive. _13, the average value _I 13 in _{g 23, g 33, I 23} , I 33 is negative. Therefore, it can be estimated that the hue change vector is oriented in the negative direction of the x-axis.

In the process of FIG. 6, the following process is further executed in order to specify the hue change vector. First, in ST 106, using the average value _{I ij} calculated in the ST105, the amount of change in hue in the x axis direction and y axis direction Cx, calculates the Cy. This calculation process is performed by calculation using a Sobel filter used for edge extraction processing of a grayscale image. Specifically, the following formulas (8) and (9) are used.

Cx = (I ₁₃ + 2I ₂₃ + I ₃₃ ) − (I ₁₁ + 2I ₂₁ + I ₃₁ ) (8)
Cy = (I ₁₁ + 2I ₁₂ + I ₁₃ ) − (I ₃₁ + 2I ₃₂ + I ₃₃ ) (9)

  The above-described change amounts Cx and Cy can be considered as the direction in which the hue data increases and the change amount in the inspection region are decomposed into the x and y axes. Therefore, in ST107, a combination of the vectors indicated by Cx and Cy is used as a hue change vector, and the length and direction of this vector are calculated from Cx and Cy.

In ST107, processing for calculating the length and direction of the combined vector of the two vectors represented by Cx and Cy is executed. In this embodiment, the direction of the combined vector is defined as an angle θ formed by the vector with respect to the positive direction of the x axis, and any one of the following formulas (A) to (E) is set according to the values of Cx and Cy. It is made to calculate using. Moreover, the length of a vector is calculated | required by (F) Formula.
When Cx> 0 and Cy ≧ 0
θ = tan ⁻¹ (Cy / Cx) (A)
When Cx> 0 and Cy <0
θ = tan ⁻¹ (Cy / Cx) +360 (B)
When Cx <0 θ = tan ⁻¹ (Cy / Cx) +180 (C)
When Cx = 0 and Cy> 0, θ = 0 (D)
When Cx = 0 and Cy <0, θ = 180 (E)

  In the case of the average value calculation example shown in FIG. 13, Cx = −17.11 and Cy = 0.07 from the above equations (8) and (9). That is, the change in hue from the red side to the blue side is represented by a vector pointing in the negative direction (left side in the figure) in the x-axis direction, and a positive direction (upper side in the figure) in the y-axis direction. It can be thought of as being represented by a facing vector.

Further, when ST107 is executed using the above-described values of Cx and Cy, (C) is applied to the calculation of the angle θ of the combined vector, and θ = 179.77 °. The vector length D is D = 17.11 according to the equation (F).
According to this calculation result, as described above, the hue change vector is oriented in the negative direction of the x axis.

Since the inspection area 31 shown in FIG. 7 is set in the land window W11 on the right side of the part, the desirable direction of the hue change vector is the direction P _A shown in FIG. Direction. According to the above calculation result, the direction of the hue change vectors extracted from the inspection region 31 of the illustrated complies with the desired direction P _A, in the inspection region 31 is considered to fillet a good condition exists be able to.

  In the fillet inspection of this embodiment, the inspection areas 31 and 32 are set in the land windows W11 and W12, respectively. After the above-described hue change vector extraction processing is performed for each of the inspection areas 31 and 32, the extraction is performed. The direction of the vector is compared with a pre-registered reference direction. However, if the magnitude of the hue change vector extracted in any inspection area is smaller than the predetermined threshold value, no hue change from red to blue appears in the inspection area. It is considered. In other words, the inspection area does not include an inclined surface corresponding to the fillet.

  FIG. 14 shows a teaching example of the reference direction of the hue change vector. In this example, the operator moves the cursor from the edge of the land toward the center C of the component while visually confirming the direction in which the color changes in the order of red, green, and blue. By doing so, the direction of movement of the cursor (indicated by arrows f and k in the figure) is designated as the reference direction. The designated directions f and k are converted into angles with respect to the positive direction of the x axis, respectively, and are registered in the teaching table 19.

  FIG. 15 shows an inspection procedure in the substrate inspection apparatus. Prior to the execution of this procedure, the clerk performs an operation of inputting the board name of the board 1T to be inspected using the input unit 22, and the determination file corresponding to the board 1T to be read is read accordingly. Is set in the memory 13 and the procedure of FIG. 15 is started. This procedure is performed for one board to be inspected 1T, and is repeated according to the number of boards to be inspected 1T. In FIG. 15, the start of the process is ST1.

  In ST1, the inspected substrate 1T is carried into the Y-axis table unit 7, and imaging by the imaging unit 3 is started. Next, the control unit 11 aligns the imaging unit 3 and the light projecting unit 4 with the substrate to be inspected 1T based on the positioning data in the determination file, and generates an image of the substrate 1T. Then, inspection windows W11, W12, W2, and W3 are set for each component on the generated image.

  In the next ST3, for the first part to be inspected, a process of extracting lands on both sides in the reference window W3 is performed, and the positions of the land windows W11 and W12 are corrected according to the extraction result. In the next ST4, it is checked whether or not to perform a fillet inspection for this part. If this determination is “YES”, then ST5 to ST9 are executed in order.

  Note that ST5 to 9 are individually performed for each of the two land windows W11 and W12. However, for the sake of simplicity, only the land window W11 will be described below as a processing target. Also in this description, it is assumed that the horizontal width direction of the component is along the x-axis direction.

  First, in ST5, processing for extracting a boundary position between the electrode of the component and the fillet is executed in the land window W11. For this processing, the determination file is set in advance for each of the binarization threshold values (R, G, B) for extracting the color of the electrode image (generally a color with a strong redness). Or the number of pixels corresponding to the vertical length of the electrode is registered (these data can also be diverted from the library data). In ST5, after the image in the land window W11 is binarized by the binarization threshold value, the direction from the inner side (part side) to the outer side of the land window W11 is set as the search direction, and the line is orthogonal to the search direction. The lines in which the colors of the electrodes are continuous for the number of pixels corresponding to the length of the vertical width are extracted. Here, the x-coordinate of the search position when the line where the color of the electrode appears is shifted from the extracted state to the non-extracted state is recognized as the boundary position between the electrode and the fillet.

  When the boundary position between the electrode and the fillet is thus extracted, in the next ST6, the inspection region 31 is set in a range from the extracted boundary position to the outer edge of the land window W11. In subsequent ST7, the hue data H1 is obtained for each pixel in the set inspection region 31 using the equations (1) to (7). In ST8, a process for extracting a hue change vector in the inspection region 31 is executed using the hue data H1 for each pixel. In ST8, the series of processes shown in FIG. 6 is executed, and an angle θ and a length D representing the direction of the hue change vector are calculated.

In the next ST9, it is determined whether or not a fillet is formed and the quality of the inclined surface of the fillet using the extracted hue change vector.
In this discrimination process, first, the length D of the hue change vector is compared with a predetermined threshold value (in the case of the formula (F), it may be about 10). In this comparison, if the length D is equal to or greater than a threshold value, the angle θ is compared with a reference angle registered in advance. If the difference of the angle θ with respect to the reference angle is within a predetermined threshold (for example, 10 °), it is determined that the fillet is well formed. On the other hand, if the angle difference exceeds the threshold value, it is determined that the tilted state of the fillet is inappropriate. If the vector length D is less than the threshold value, it is determined that the inspection area 31 does not include an inclined surface corresponding to a fillet.

  When the fillet inspection is completed in this way, the process proceeds to ST10, and other inspections (such as the presence / absence of parts and positional deviation) are performed. Since this inspection is the same as the conventional one, the details are omitted.

  Thereafter, in the same manner, the inspection for each part on the image is sequentially executed. The determination result for each component is temporarily stored in the memory 13. When the inspection for one screen is completed, the X-axis table unit 6 and the Y-axis table unit 7 are moved to change the field of view of the imaging unit 3, and imaging is performed again. A similar test is performed.

  When the processing for all the parts is completed, ST11 becomes “YES”, and the process proceeds to ST12. Based on the determination result stored in the memory 13, it is determined whether or not there is a part determined to be defective. If it is determined that there is a defective portion, the process proceeds to ST13, where it is determined that the inspected substrate 1T is defective. On the other hand, if the good determination is obtained for any of the components, the process proceeds to ST14 and it is determined that the inspected substrate 1T is a good product. Thereafter, the process proceeds to ST15, the determination result is output, and the process is terminated.

  In the above embodiment, at the time of teaching, the reference direction of the hue change vector is registered according to the operation of the operator. However, the present invention is not limited to this, and the land windows W11 and W12 are processed by the same processing as that at the time of inspection. May be extracted, and the extraction result may be registered. In this case, the direction and length can be compared between the hue change vector extracted at the time of inspection and the registered hue change vector.

  Further, the extraction position of the hue change vector may be registered during teaching. In this case, it is possible to detect the misalignment of the fillet by comparing the extracted position of the hue change vector at the time of inspection with the registered position.

Further, according to the color change vector, when a plurality of identical components are mounted on the board, if the reference direction of the color change vector is registered for one of the components, the registered data is transferred to other components. can do. As described above, even if the parts are the same, the inclination angle of the fillet may be different depending on the peripheral mounting density, but the direction of the hue change vector does not change. Therefore, even if there is a variation in the inclination angle of the fillet between the same parts, it is possible to use common inspection information and improve the efficiency in teaching. Also, for a fillet that is steep due to the small land and only the blue region is extracted in the binarization process, the direction of the gradient change can be extracted with the hue change vector. Inspection accuracy can be greatly increased.
The reference direction of the hue change vector can be registered as library data.

  Further, the hue data H1 used in the above embodiment indicates which position in the range from red to blue of the hue circle the color of the pixel of interest corresponds to, but instead of this, for example, blue For the gradation data B corresponding to the light source, the hue data b according to the following equation (10) may be obtained.

  The hue data b according to the above equation (10) indicates the ratio of the blue component of the color components that determine the brightness of the pixel. Similarly to the hue data H1, the hue data b increases as the bluish color increases and decreases as the redness increases. Therefore, a hue change vector is set in the same manner as in the previous embodiment, The inclination direction of the fillet can be determined.

Furthermore, in this embodiment, the color images are processed using the light sources 8, 9, and 10 that emit different color lights. Instead, these light sources 8, 9, and 10 are used as white light sources. Even when imaging with a monochrome camera, it is possible to execute the same processing as in the above-described embodiment.
For example, the light source 8 is first turned on to take an image, and the obtained grayscale image data is regarded as R gradation data and temporarily stored in the memory 13. Next, the light source 9 is turned on for imaging, and the obtained grayscale image data is regarded as G gradation data and temporarily stored in the memory 13. Finally, the light source 10 is turned on to take an image, and the obtained grayscale image data is regarded as B gradation data and temporarily stored in the memory 13. In this way, the three shaded image data acquired in sequence are handled as R, G, and B gradation data, respectively, and the hue change vector extraction process is executed by applying the above-described embodiment. According to this method, the time required for generating the image data becomes long, but thereafter, the fillet inspection can be executed by the same algorithm as that in the case of using the color image.

It is a block diagram which shows the structure of the board | substrate inspection apparatus to which this invention was applied. It is a figure which shows the other structural example of a light projection part. It is a figure which shows the example of a setting of a window. It is a figure which shows the example of a setting of the test | inspection area | region for a fillet test | inspection. It is a figure which shows the distribution state of the color in the fillet on an image corresponding to the side surface shape. It is a flowchart which shows the procedure of the extraction process of a hue change vector. It is a figure which shows the example of calculation of the hue data in a test | inspection area | region. It is a figure which shows the structure of the mask scanned to a test | inspection area | region. It is a figure which shows the content of the data extracted from the test | inspection area | region by the scan of the mask. It is the figure which matched the extraction data obtained by the 1st line scan, and the arrangement | sequence data of the variation | change_quantity of a hue. It is the figure which matched the extraction data obtained by the scanning of the 2nd row, and arrangement data of the amount of change of hue. It is the figure which matched the extraction data obtained by the scanning of the 3rd row, and the arrangement data of the amount of change of hue. It is a figure which shows the arrangement | sequence data of the variation | change_quantity of the hue obtained in each scanning position, and the calculation result of an average value. It is a figure which shows an example of the method of teaching the reference | standard direction of a hue change vector. It is a flowchart which shows the procedure of a test | inspection. It is a figure which shows the structure of the optical system of the conventional board | substrate inspection apparatus. It is a figure which shows the principle of the solder recognition process by the optical system of FIG. It is a figure which shows the principle of the solder recognition process by the optical system of FIG.

Explanation of symbols

1T Board to be inspected 3 Imaging unit 4 Projecting unit 5 Control processing unit 8, 9, 10 Light source 11 Control unit 15 Image processing unit 18 Inspection unit 19 Teaching table 23 CRT display unit 25 Transmission / reception unit 26 External memory device

Claims (10)

A plurality of light sources that emit different colored lights are arranged and illuminated with respect to an inspection object having an inclined surface on the surface so that elevation angles viewed from the arrangement position of the inspection object are different, and above the arrangement position In a method for inspecting the surface state of the inspection object using an image obtained by an installed imaging means,
A first step of setting an inspection region including an inspection object in the image, and extracting a direction in which a hue change corresponding to the arrangement of each light source occurs in the inspection region, and extracting in the first step A second step of comparing the determined direction with a preset reference direction, and a third step of determining the quality of the surface state of the inspection object based on the comparison result of the second step A surface condition inspection method characterized by: The plurality of light sources are arranged so that the respective emission colors are arranged along the direction of hue change on the hue ring with respect to the inspection object,
In the first step, for each pixel in the inspection area, one-dimensional hue data whose value changes along the hue circle is acquired, and the hue data changes to a large value or changes to a small value. The surface state inspection method according to claim 1, wherein a direction in which to perform the extraction is extracted.   In the first step, for each pixel in the inspection area, one-dimensional hue data reflecting the intensity level of the color component corresponding to any one of the plurality of light sources is acquired, and the hue data 2. The surface condition inspection method according to claim 1, wherein a direction in which the value changes to a large value or a direction in which the value changes to a small value is extracted.   In the first step, while paying attention to a plurality of pixels in the inspection region in order, extracting a change amount of hue data for each of a plurality of directions passing through the pixel of interest; and hue data extracted for each pixel of interest The surface state inspection method according to claim 2 or 3, wherein a step of setting a plurality of vectors by averaging the amount of change for each direction and a step of setting a combined vector of these vectors are executed. Illumination means comprising a plurality of light sources that emit different color lights and arranged such that the elevation angles when viewed from the arrangement position of the inspection object are different,
Imaging means arranged above the position of the inspection object;
An image input means for capturing an image generated by the imaging means in a state where each light source of the illumination means is turned on;
An extraction means for setting an inspection area including an inspection object in the image input by the image input means, and extracting a direction in which a hue change corresponding to the arrangement of each light source occurs in the inspection area;
Registration means for registering a reference direction extracted by the extraction means when the surface condition of the inspection object is good;
A determination unit that compares the direction of change extracted by the extraction unit with a reference direction registered by the registration unit to determine whether the surface state of the inspection object is good;
A surface condition inspection apparatus comprising output means for outputting a discrimination result by the discrimination means. Each light source of the illuminating means is arranged so that each emission color is aligned along the direction of the hue change on the hue ring with respect to the inspection object,
The extraction unit acquires one-dimensional hue data whose value changes along the hue circle for each pixel in the inspection area, and the direction in which the hue data changes to a large value or a small value. The surface condition inspection device according to claim 5, wherein the surface condition inspection device is extracted.   The extraction means acquires one-dimensional hue data reflecting the degree of intensity of a color component corresponding to one of the plurality of light sources for each pixel in the inspection region, and the hue data is large. The surface condition inspection apparatus according to claim 5, wherein a direction that changes to a value or a direction that changes to a small value is extracted.   The extraction means is a means for extracting a change amount of hue data for each of a plurality of directions passing through the pixel of interest while paying attention to a plurality of pixels in the inspection region in order, and a change of the hue data extracted for each pixel of interest. 8. The surface condition inspection apparatus according to claim 6, comprising means for setting a plurality of vectors by averaging the quantities for each direction, and means for setting a combined vector of these vectors. An apparatus for inspecting a board on which a plurality of components are soldered,
Illumination means comprising a plurality of light sources that emit different color lights arranged so that the elevation angles when viewed from the substrate to be inspected are different,
Imaging means disposed above the substrate;
An image input means for capturing an image generated by the imaging means in a state where each light source of the illumination means is turned on;
Extraction means for setting an inspection region at each soldering site in the image input by the image input means, and extracting a direction in which a hue change corresponding to the arrangement of each light source occurs for each of the inspection regions;
Registration means for registering the reference direction extracted by the extraction means when the surface condition of the solder in the inspection area is good;
A determination unit that performs a process of comparing the direction of change extracted by the extraction unit with a reference direction registered by the registration unit for each inspection region, and determines whether each soldering part is good or bad;
A substrate inspection apparatus comprising: output means for outputting a discrimination result by the discrimination means.   When a plurality of components of the same type are mounted on the board to be inspected, the registration means applies the reference direction registered for one soldering part of these components to other components of the same type. The board inspection apparatus according to claim 9.