JP2002049916A - Device and method for extracting color - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a color extracting device and a color extracting method capable of appropriately and efficiently performing color extraction in color image processing. SOLUTION: In color extraction processing that extracts a prescribed color from a color image, the respective pieces of element data H, S and I of HIS image data obtained by converting RGB image data acquired by imaging into a color model are converted by using a conversion table 7c where numerical data stored in a library part 7d are respectively read and set, and a synthetic image is formed on the basis of converted data. Thus, color extraction can be performed with high accuracy in a color image processing, a histogram showing the distribution state of image information is further prepared and displayed in each element data, and a conversion table provided with an appropriate characteristic can quickly be set with high operability by overlapping the display and the table 7c and confirming the characteristic of the conversion table.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a color extracting apparatus and a color extracting method for extracting a specific color in a color image in color image processing.

[0002]

2. Description of the Related Art A method using color image processing is known as an appearance inspection method performed on a mounting board on which electronic components are mounted. In this inspection method, a color image obtained by imaging the inspection target is subjected to image processing to separate and detect an inspection target portion such as an electronic component. This separation and detection is performed by extracting a specific color from the color image. This is done by extraction.

[0003]

In order to extract a specific color from a color image, a dedicated filter for separation is often set according to a detection target. This separation filter is generally provided as a conversion table for converting image data, and it is necessary to appropriately set the conversion table so that the color specifying the detection target part can be clearly identified from the surrounding area. There is. However, conventionally, a method for correctly and efficiently setting a conversion table having such characteristics has not been established, and it takes time and effort to set the conversion table. There was a problem that it was difficult to perform an accurate visual inspection because it could not be performed.

Accordingly, an object of the present invention is to provide a color extraction device and a color extraction method that can appropriately and efficiently perform color extraction in color image processing.

[0005]

According to a first aspect of the present invention, there is provided a color extracting apparatus for extracting a predetermined color from a color image, wherein the first color data is stored in accordance with the first numerical data. A first conversion table for converting first element data forming the image information of the color image; and a second element storing second numerical data and forming the image information of the color image according to the second numerical data. A second conversion table for converting data, a third conversion table for storing third numerical data, and converting third element data constituting image information of the color image according to the third numerical data; And a synthesizing unit for synthesizing the three types of data respectively converted by the first, second and third conversion tables to form a synthesized image.

According to a second aspect of the present invention, in the color extracting apparatus of the first aspect, the first, second, and third numerical data correspond to a color to be extracted.

According to a third aspect of the present invention, there is provided the color extracting apparatus according to the second aspect, wherein the first, second, and third numerical data are set as one set, and the set of data is set to a plurality of different sets. And a library unit storing a plurality of sets for each of the colors, and selecting one set from the plurality of sets stored in the library unit, and assigning the set of the first, second, and third numerical data to the first, second, and third, respectively. , A selection means for storing the data in the third conversion table.

According to a fourth aspect of the present invention, there is provided a color extracting apparatus according to the first aspect, wherein a display section for displaying a color image and a specific area of the color image displayed on the display section are sampled. A histogram creation unit that creates a histogram for each component of color image information obtained by converting a color image of the sampling area by a color model conversion unit and displays the histogram on the display unit; A conversion table display processing unit that displays a geometric figure based on the first, second, and third numerical data on the display unit; and an operation unit that deforms the geometric figure displayed on the display unit. The first, second, and third numerical data are changed based on the geometrical figure.

According to a fifth aspect of the present invention, in the color extracting apparatus of the first aspect, the display unit displays the histogram and the geometrical figure in a superimposed manner.

According to a sixth aspect of the present invention, there is provided a color extracting method for extracting a predetermined color from a color image, wherein the first element data constituting the image information of the color image is converted according to first numerical data. A first conversion step, a second conversion step of converting second element data constituting image information of the color image according to second numerical data, and an image information of the color image according to third numerical data. The method includes a third conversion step of converting the constituent third element data and a synthesis processing step of synthesizing the data converted in the first, second, and third conversion steps to form a synthesized image.

The color extracting method according to claim 7 is the color extracting method according to claim 6, wherein the first, second, and third numerical data correspond to the color to be extracted.

The color extracting method according to claim 8 is the color extracting method according to claim 6, wherein a display step of displaying a color image and a specific region of the displayed color image are designated as a sampling region. A sampling area designation step, a histogram creation step of creating a histogram for each component of color image information obtained by converting a color image of the sampling area by a color model conversion unit and displaying the histogram on a display unit, (2) a conversion table display processing step of displaying a geometrical figure based on the third numerical data on the display unit, and transforming the geometrical figure displayed on the display unit with the operation unit, Changing the first, second, and third numerical data based on the numerical data.

A color extracting method according to a ninth aspect is the color extracting method according to the eighth aspect, wherein the display unit displays the histogram and the geometric figure in a superimposed manner.

A storage medium according to a tenth aspect of the present invention is a storage medium storing a color extraction processing program for extracting a predetermined color from a color image, wherein the image information of the color image is configured according to first numerical data. A first conversion processing for converting one element data, a second conversion processing for converting second element data constituting image information of the color image according to second numerical data, and the color processing according to third numerical data A third conversion process for converting third element data constituting the image information of the image, and a synthesis process for forming a synthesized image by synthesizing the data converted by the first, second and third conversion processes The processing program including is stored.

[0015] The storage medium according to claim 11 is the storage medium according to claim 10.
A display process for displaying a color image, a sampling region specifying process for specifying a specific region of the displayed color image as a sampling region, and a color model conversion unit for converting the color image of the sampling region into a color model. A histogram creation process for creating a histogram for each component of the color image information obtained by the conversion and displaying the histogram on a display unit, and a geometric figure based on the first, second, and third numerical data is displayed on the display unit. And a transformation table display process to be displayed on the display unit, and transforming the geometrical figure displayed on the display unit by the operation unit, and
A processing program including a change process for changing the second and third numerical data is stored.

According to the present invention, in a color extraction process for extracting a predetermined color from a color image, three element data constituting image information of the color image are converted using a conversion table, and based on the converted data, By forming a composite image, color extraction can be performed with high precision in color image processing. Further, a histogram indicating the distribution state of image information is created and displayed for each element data, and this histogram and a conversion table By checking the characteristics of the conversion table by superimposing them, a conversion table having appropriate characteristics can be set with good operability.

[0017]

Embodiments of the present invention will now be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a board appearance inspection apparatus according to an embodiment of the present invention. FIG. 2 is a functional block diagram showing a color extraction processing function of the board appearance inspection apparatus according to an embodiment of the present invention. FIG. 3 is an explanatory diagram of the color extraction processing of the board appearance inspection apparatus according to one embodiment of the present invention, and FIG. 4 is a functional block showing the color extraction setting processing function of the board appearance inspection apparatus of one embodiment of the present invention. FIGS. 5 and 5 are flowcharts of a board inspection process for an electronic component according to an embodiment of the present invention. FIGS.
FIGS. 8, 9 and 10 are views showing display screens for setting color extraction filters of the board appearance inspection apparatus according to one embodiment of the present invention.

First, the configuration of the board appearance inspection apparatus will be described with reference to FIG. In FIG. 1, an XY table 1
A substrate 2 is held thereon, and a plurality of types of different electronic components 3 are mounted on the substrate 2. Above the substrate 2, an imaging unit including the illumination unit 4 and the camera 5 is provided. The illumination unit 4 is a ring-shaped light-emitting body 4a that emits white light.
And a substantially hemispherical reflector 4b. The white light emitted from the light emitter 4a toward the inner surface of the reflector 4b is diffusely reflected by the spherical reflector to illuminate the substrate 2 below. .

A camera 5 is disposed above the illuminating section 4 in a downward direction. The camera 5 is illuminated with white diffusely reflected light from the substrate 2 through an opening provided above the reflector 4b. Is imaged. At this time, by driving the XY table 1, the board 2 moves horizontally, and an arbitrary electronic component 3 mounted on the board 2 can be positioned just below the camera 5 to capture an image.

The camera 5 includes a camera control unit 6
It is connected to the. The camera control unit 6 controls the imaging operation of the camera 5 such as the imaging timing, and extracts three primary colors of R (red), G (green), and B (blue) from the imaged data and outputs them as image data for each element. I do.

The storage section 7 comprises an image storage section 7a, a program storage section 7b, a conversion table 7c, and a library section 7d. The image storage unit 7a stores image data for each of R, G, and B elements output from the camera control unit 6, and H (hue), S (color model converted) described later.
Three-dimensional HSI image data in which each element of (saturation) and I (brightness) is one independent dimension,
Various image data such as composite image data obtained by performing data conversion processing on SI image data and the like are stored.

The program storage unit 7b stores various processing / operation programs such as a color model conversion program, a board inspection program, a color extraction processing program, a color extraction setting program, and an overall processing program. The conversion table 7c includes three element data constituting the image information of the color image obtained by performing the color model conversion of the RGB image data, that is, H which is the first element data.
(Hue), S (saturation) which is the second element data, and I (brightness) which is the third element data. The conversion table is a data conversion table for filtering. As described later, first, second, and third numerical data set for each element data are stored. The library unit 7d stores the numerical data set in the conversion table 7c for each type of color to be extracted.

The calculation unit 8 is a CPU, and performs calculations and processes according to various programs stored in the program storage unit 7b. The display unit 9 is a monitor, which displays a guidance screen for operation and input, and also displays an image captured by the camera 5 and converted into a color model. Operation / input unit 1
Reference numeral 0 denotes input means such as a keyboard, a mouse, a touch panel, etc., for inputting operation commands and various data.

The storage medium drive device 11 is an external data reading device such as a floppy (registered trademark) disk drive. When the various processing programs stored in the program storage unit 7b are provided by being stored in the storage medium 11a such as a floppy disk, the storage medium drive device 11 reads the various processing programs from the storage medium 11a. Thus, a general personal computer having an input function, an arithmetic mechanism, a storage function, and a display function can be combined with an imaging unit having a color image capturing function to function as a color extracting device.

Next, with reference to FIGS. 2 and 3, a description will be given of a color extraction process in a board appearance inspection. 2 and 3, an RGB image storage unit 13, an HSI image storage unit 14,
The composite image storage unit 18 corresponds to the image storage unit 7a shown in FIG. The color model conversion unit 14 and the synthesis processing unit 17 show processing functions performed by the calculation unit 8 shown in FIG.

The RGB image data of the substrate 2 to be imaged, which is captured by the camera 5 and output by the camera control unit 6, is stored in the image storage unit 7a. The RGB image data is converted into HSI image data by the color model conversion unit 14, and the converted HSI image data is stored in the HSI image storage unit 15. Then, a color extraction process for extracting a specific color based on the HSI image data is performed.

Here, as shown in FIG. 3, the purpose is to inspect the mounting state of the electronic component 3 (square capacitor component) mounted on the substrate 2, that is, the presence or absence of the electronic component 3 and the posture in the mounting state. An example is shown in which the upper surface of the electronic component 3 is set as a target region for color extraction, and this region is converted into a monochrome image in a conversion form that can be clearly separated from other regions. By performing such processing, RGB
Even when the separation is difficult on the image, the target area can be clearly separated from the surroundings and taken out, and the various quantities such as shape recognition and dimensions can be easily calculated.

As shown in FIG. 3, the image of the substrate 2 including the electronic component 3 captured by the camera 5 is subjected to color model conversion from RGB image data to HSI image data. Here, a process of converting color data represented in RGB dimensions into data in HSI dimensions is performed for each pixel constituting an image, and H,
Data is separated for each element of S and I, and
It is output as element data, second element data, and third element data.

Thus, the HSI image storage unit 15 stores
An image in which the data of H, S, and I correspond to each pixel, that is, a color conversion image 1 represented only by H (hue)
5a, a color conversion image 15b expressed only by S (saturation) and a color conversion image 15c expressed only by I (lightness) are stored. Then, data conversion using the conversion table 7c (the first conversion table 16a, the second conversion table 16b, and the third conversion table 16c) is performed on these image data.

This data conversion is performed by replacing the numerical values of H, S, and I obtained for each pixel with new numerical values according to a predetermined conversion rule. This transformation law is
The first conversion table 16a, the second conversion table 16b, and the third conversion table 16c are provided in the form of a data conversion table composed of numerical data. 3 are stored, and the respective element data are converted according to the numerical data.

These three numerical data form a set,
One set corresponds to a specific color to be extracted. That is, these first, second, and third numerical data are unique data corresponding to the color to be extracted. The library unit 7d stores a plurality of sets of numerical data for a plurality of different colors, and specifies a color to be extracted.
One set of numerical data corresponding to the color is read from the library unit 7d by the board inspection program 19.

By storing the read first, second, and third numerical data in the corresponding conversion tables, the setting of the conversion table for extracting the color is automatically performed. That is, the board inspection program 19 selects one set from the plurality of sets of data stored in the library unit 7d, and converts the set of first, second, and third numerical data into first, second, and third data, respectively. 3 is a selection means to be stored in the conversion table.

Here, a specific example of the conversion table 7c will be described. The graphs of the first, second, and third conversion tables 16a, 16b, and 16c shown in FIG. 3 show the characteristics of the conversion rule. Taking the first conversion table 16a as an example,
A graph showing the characteristics of the conversion law will be described. The first conversion table 16a shows the correspondence between the input value H and the output value H 'of the attribute value indicating the hue, and each numerical value H (input value) on the horizontal axis represents the characteristic shown on this graph. It corresponds to each numerical value H ′ (output value) on the vertical axis via a line.

The characteristic line indicating the conversion characteristic of the first conversion table 16a is represented by a trapezoidal geometric figure and has the following characteristics. That is, if the numerical value of H on the horizontal axis is out of the range of the lower base of this trapezoid, the corresponding numerical value of H 'is 0 (zero).
Becomes If the numerical value of H is within the range of the upper base of the trapezoid, the corresponding numerical value of H ′ is the maximum value Hmax. And the area within the hypotenuse of the trapezoid is a fuzzy area,
As shown in FIG. 3, the numerical values H in these ranges are converted into corresponding numerical values H 'on the vertical axis by the oblique sides of the trapezoid. The second conversion table 1 is also used for S and I data.
6b and the third conversion table 16c,
The image is converted to an S ′, I ′ image.

The synthesizing unit 17 synthesizes three types of data converted by the first, second, and third conversion tables to form a synthesized image. That is, the data-converted H ′, S ′, and I ′ images are combined by the combining processing unit 17 to form a color extracted image. The synthesized color extraction image is stored in the synthesized image storage unit 18. In FIG. 3, a portion of the electronic component 3 to be detected is a white portion 18b, and the other portion is a color extracted image (monochrome image) 18 converted into a non-white portion 18a whose color changes from black to gray.
An example of obtaining A is shown.

In the case of performing such a color extraction, a determination condition such that only pixels whose H, S, and I values after conversion by the conversion table 7c satisfy all predetermined conditions is extracted as a white portion. May be set. That is, if any one of the numerical values of H, S, and I of a certain pixel is out of a determination condition such as a threshold value set as a color extraction range, this pixel is excluded from a color extraction target. Therefore, it indicates that a subtle difference in attribute can be sharply identified, and high-precision separation can be performed even when it is difficult to separate regions on the RGB image data.

Next, the color extraction setting function of the board appearance inspection apparatus will be described with reference to FIG. In this color extraction setting, the characteristics of the conversion table 7c are set so that appropriate color extraction can be performed for each area of the inspection target in the HSI image obtained by converting the RGB image data of the substrate 2 to be inspected into a color model. Things.

In FIG. 4, a sampling area designating section 2
1 performs a process of designating a specific area to be a sampling area for color extraction on the display screen as a sampling area by an input operation from the operation unit 22. When the sampling area is specified, the histogram creating unit 20 reads H, S, and I image data in the area specified as the sampling area from the image data stored in the HSI image storage unit 15 and A histogram showing the distribution state of the image data is created and displayed on the monitor of the display unit 9. Thereby, the attribute data (here, three types of data of H, S, and I) indicating the characteristics of the material and the surface properties of the area are quantified on the image data obtained by imaging the area, The distribution state is determined.

The numerical data storage unit for editing 23 reads, from the library unit 7d, numerical data to be edited, that is, numerical data defining the characteristics of the conversion table corresponding to the color to be extracted. The conversion table display processing unit 24 displays a predetermined geometric figure formed by characteristic lines indicating the characteristics of the conversion table based on the numerical data in the numerical data storage unit 23 to be edited according to an instruction from the operation unit 22. To be displayed.

At this time, this geometrical figure is displayed so as to be superimposed on the histogram indicating the distribution state of the image elements as described later. This makes it possible to visually compare the characteristics of the conversion table with the actual distribution state of the attribute values in the color extraction target area, and easily determine whether the characteristics of the conversion table are appropriate.

In the present embodiment, the geometric figure is set to a trapezoidal shape including the above-mentioned histogram. The geometric figure can be corrected and deformed on the display screen by an input operation from the operation unit 22. By this correction and deformation operation, the numerical data is automatically corrected based on the deformed geometric figure,
The data stored in the library unit 7d is similarly modified. As the geometric figure, a simple step-like figure that does not include a fuzzy region may be used in addition to the trapezoidal shape.

The board appearance inspection apparatus is configured as described above. Next, the board appearance inspection performed by the board inspection apparatus will be described with reference to FIG. First, an image is captured by the camera 5, whereby RGB image data of the inspection target site is captured (ST1).
Next, numerical data of the color to be extracted is set in the first, second, and third conversion tables (ST2). That is, by reading the numerical data corresponding to the color of the part to be inspected and storing it in the conversion table 7c, the conversion table 7c
Is set to the conversion characteristic corresponding to the color.

Next, the read RGB image data is
The color model conversion unit 14 converts the color model into HSI image data (ST3). After that, data conversion is performed for each element of the HSI image data by the first, second, and third conversion tables, and then color extraction processing is performed by performing image synthesis processing. That is, a first conversion step of converting first element data forming image information of a color image according to first numerical data, and a second conversion step of converting second element data forming image information of a color image according to second numerical data. In the second conversion step, the third conversion step of converting the third element data constituting the image information of the color image in accordance with the third numerical data, and the first, second and third conversion steps. A combining processing step of combining the converted data to form a combined image is performed. Here, if the region includes a plurality of colors to be extracted, (S
The processing of T2) to (ST4) is repeated the required number of times.

Then, a judgment process is performed using the synthesized image thus synthesized (ST5). In other words, by processing image information for a predetermined area in the composite image, various inspection quantities (area size, area, center of gravity position, boundary line, etc.) corresponding to each inspection item are obtained, and these inspection various quantities are obtained. Is compared with a threshold value set for each item. Thus, one step of the appearance inspection of the mounting board is completed.

Next, the color extraction setting processing will be described with reference to FIGS. First, as shown in FIG. 6, an RGB image of the substrate 2 captured by the camera 5 is displayed on the display screen 30 (display step). Here, the electronic component 3 to be inspected is displayed in a state of being soldered to the land 2a of the substrate 2. Next, a specific area in the displayed RGB image is designated as a sampling area (sampling area designating step). Various touch panel buttons are set in the display screen 30. By operating the sampling area specification button 31, an area frame 3a for specifying an area to be sampled appears. Inching button 3
4, the screen is moved, and by operating an arrow on the screen with a mouse, an area frame 3 in the screen 30 is displayed.
The position and size of “a” can be set.

FIG. 6 shows a state in which the area frame 3a is positioned on the upper surface of the electronic component 3. In this state, the area is specified by operating the sampling area specifying button 31, and the cancel button 35 is operated. Then, the area designation is canceled. By operating the histogram display button 32, a screen shown in FIG. 7 is displayed. By operating the conversion table edit button 33, a screen shown in FIG. 9 is displayed so that a conversion table editing operation described later can be performed.

Next, a histogram is created and displayed for each of the items H, S, and I, which are the components of the color image information (histogram creation step). In FIG. 7, a display screen 30 shows a distribution diagram of numerical data of H, S, and I in the area obtained from image data in the sampling area, that is, H, S obtained for each pixel constituting the area. , I, the distribution chart 35a, 35b, which represents the cumulative distribution of the numerical data for each item of H, S, I.
35c is displayed. In these distribution charts, the horizontal axis represents the values of H, S, and I, and the vertical axis represents the cumulative frequency (the number of pixels) of the data at each value. , I, that is, histograms 36a, 36b, 36c representing the characteristics of this color are displayed.

Then, the histograms 36a, 36b, 3
6c, a geometric figure showing the characteristics of the conversion table is displayed based on the numerical data read from the numerical data storage unit 23 to be edited. In this case, if no numerical data is stored in the edit target numerical data storage unit 23, the displayed histograms 36a, 36b, and 36c are displayed.
Geometric figures 37a and 37 that include the mountains of
b and 37c are automatically displayed, and numerical data corresponding to these geometric figures 37a, 37b and 37c are provisionally set and stored in the edit target numerical data storage unit 23.

On the screen 30, together with these displays, a reference frame (color palette) for visually referring to the hue and saturation ranges covered by the conversion table set here.
38 and a reference frame 39 for visually referring to the brightness range covered by the conversion table.

The reference frame 38 displays the color change two-dimensionally by changing the hue H in the horizontal direction and the saturation S in the vertical direction. The reference frame 39 displays the change in brightness by changing the brightness in the left-right direction. The windows 38a, 39a shown in the reference frames 38, 39
The hue / saturation and lightness in the area can be visually confirmed by the position. In the reference frame 40, an image indicating the color of the color capturing area is displayed.

By operating the color extraction button 44, the numerical data in the numerical data storage unit for editing 23 is stored in the first, second, and third conversion tables 16a, 16b, and 16 respectively.
c and the screen of the color extraction image shown in FIG. 8 is displayed. This makes it possible to confirm whether or not the intended color extraction has been performed. That is, if a monochrome image in which the region 46b corresponding to the upper surface of the electronic component 3 is extracted as a white image and the other regions 46a are displayed between black and gray is obtained, the expected color extraction has been performed. become.

When the target conversion table is obtained in this way, registration processing is performed. That is, by operating the registration button 41 on the screen 30, the numerical data stored in the numerical data storage unit for editing 23 is registered in the library unit 7d. If the numerical data to be registered is new data other than the registered colors, a name or code for specifying the numerical data is added.
By operating the correction button 42, the screen shown in FIG. 9 is displayed, and the geometric figure showing the characteristics of the conversion table 7c can be changed on the screen. Also, a cancel button 43
By operating, the setting operation on this screen is canceled, and the screen returns to the screen shown in FIG.

When inputting new numerical data or performing editing work for creating a conversion table by diverting registered data, the user operates the conversion table edit button 33 shown in FIG. The edit screen shown in is displayed. In FIG. 9, a numerical data name display frame 51 is displayed on the screen 30, and a plurality of sets of numerical data (first, second, and second) corresponding to each color type stored in the library unit 7d are displayed.
The names of the third three are made into a list.

By operating the switch 51a to highlight the specific numerical data in the list, the numerical data is read from the library unit 7d. Then, geometric figures 37a, 37b, and 37c indicating the conversion table 7c corresponding to these numerical data are displayed, and these numerical data are stored in the edit target numerical data storage unit 23. By operating the enter button 50, the screen returns to the histogram display screen shown in FIG.

When the comparison sample button 45 is operated, the display screen is switched to the screen shown in FIG. 6, and the comparison sample operation can be performed in this state. That is, when the same sampling is performed on an area other than the target sampling area on the screen and the above conversion table is applied to this image, the same color extraction result as the color extraction image shown in FIG. Is obtained, it indicates that these two areas cannot be separated by the set conversion table.

Therefore, in such a case, it is necessary to correct the conversion table. This correction process is performed by operating the correction button 42 on the screen 30 shown in FIG. By this button operation, a correction screen shown in FIG. 10 is displayed. Here, distribution diagrams 35a, 35b, 35
Geometric figure 3 defining the conversion table set on c
A process of deforming the shapes of 7a, 37b, and 37c and changing the corresponding numerical data based on the changed geometric shape is performed (numerical data changing step).

FIG. 10 shows a geometrical figure 37b among these.
The example which deform | transforms the shape of FIG. In other words, a deformation target is specified by touching a break point of the geometrical figure to be deformed, and the break point is highlighted on the screen. By operating the move button 48,
The broken point of the geometrical figure 37b to be deformed is moved in a desired direction. As a result, the geometrical figure defining the conversion table is deformed, the numerical data is changed based on the geometrical shape after the deformation, and the reference frame 3 is changed with this deformation.
The display window 38a in 8 also moves according to the deformation.
Thereafter, the above-described color extraction is again performed, and if it is confirmed that the area is clearly separated from other areas, the numerical data corresponding to the conversion table 7c is stored in the library unit 7d.

As described above, the following effects can be obtained by setting the conversion table so as to include these distribution ranges based on the distribution state of the numerical data of H, S, and I obtained by actual imaging. be able to. That is,
Materials and surface properties and imaging conditions of electronic components imaged at the time of inspection do not always maintain uniformity with high accuracy. Therefore, H, which is obtained by performing color image conversion of image data of an electronic component including such variations, is obtained.
The data of S and I also vary within a certain width distribution range.

Therefore, by performing data conversion using the conversion table set as described above, the H, S, I data obtained by imaging due to variations in the material and surface properties of the electronic parts, etc. Even if there is some variation, as long as these data are located within a preset distribution range, they are determined as data having an attribute unique to the color capture area.

As described above, the present invention converts RGB image data into HSI image data in a board appearance inspection, and uses the HSI image data as a target of color extraction processing. As a result, as described above, if even one of the three elements has a difference, the region is clearly separated from the others. Therefore, even in an object that has a small difference in brightness and is difficult to separate in an RGB image. If there is a difference in hue or saturation, highly accurate color extraction becomes possible. Therefore, when performing image recognition based on the conventional RGB image data,
It is possible to reduce the number of dedicated filtering processes required depending on the target, thereby greatly improving the processing efficiency and operability.

Further, when setting the conversion table, H,
By displaying a histogram indicating the distribution state of the attribute values for each of the S and I image elements, and confirming the characteristics of the conversion table with the conversion table superimposed on the histogram, a color extraction target area of the conversion table is displayed. Can be visually and intuitively determined, and the setting of the conversion table for the color extraction setting can be quickly performed appropriately and with good operability.

In the present embodiment, the RGB image data obtained by the camera 5 is converted into HSI image data by a color model conversion, and the HSI image data is converted into a HSI image data.
Although an example in which image synthesis processing is performed is shown, the type of image data is not limited to HSI image data, and other types of element data constituting color image information may be used as processing targets. .

[0063]

According to the present invention, in a color extraction process for extracting a predetermined color from a color image, three element data constituting image information of the color image are converted using a conversion table, and the converted data is converted. Since the composite image is formed based on the color image processing, it is possible to perform color extraction with high accuracy in color image processing, and further create and display a histogram indicating a distribution state of image information for each element data,
Since the characteristics of the conversion table are checked by superimposing the conversion tables on the display, it is possible to quickly set a conversion table having appropriate characteristics with good operability.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a board appearance inspection apparatus according to an embodiment of the present invention.

FIG. 2 is a functional block diagram showing a color extraction processing function of the board appearance inspection apparatus according to the embodiment of the present invention;

FIG. 3 is an explanatory diagram of a color extraction process of the board appearance inspection apparatus according to the embodiment of the present invention;

FIG. 4 is a functional block diagram showing a color extraction setting processing function of the board appearance inspection apparatus according to the embodiment of the present invention;

FIG. 5 is a flowchart of a board inspection process for an electronic component according to an embodiment of the present invention.

FIG. 6 is a view showing a display screen of a color extraction filter setting of the board appearance inspection apparatus according to the embodiment of the present invention;

FIG. 7 is a diagram showing a display screen for setting a color extraction filter of the board appearance inspection apparatus according to the embodiment of the present invention;

FIG. 8 is a view showing a display screen of a color extraction filter setting of the board appearance inspection apparatus according to the embodiment of the present invention;

FIG. 9 is a diagram showing a display screen of a color extraction filter setting of the board appearance inspection apparatus according to the embodiment of the present invention;

FIG. 10 is a diagram showing a display screen for setting a color extraction filter of the board appearance inspection apparatus according to the embodiment of the present invention;

[Explanation of symbols]

 2 board 3 electronic component 5 camera 7 storage unit 7a image storage unit 7b program storage unit 7c conversion table 7d library unit 8 operation unit 9 display unit 10 operation / input unit 14 color model conversion unit 20 histogram creation unit 21 sampling area designation unit 22 Operation unit 24 Conversion table display processing unit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06T 1/00 305 G06T 1/00 305B F term (Reference) 2F065 AA12 AA17 AA21 AA51 AA58 CC01 CC25 FF04 GG17 GG24 JJ03 JJ09 LL19 PP12 QQ01 QQ24 QQ29 QQ31 QQ43 SS13 2G051 AA65 AB14 CA04 CB01 DA07 EA14 EA17 EC02 FA01 FA02 5B057 AA03 BA02 CA01 CA08 CA12 CA16 CB01 CB08 CB12 CB16 CC01 CE08 CE17 CH07 CH02 DA02 DB02 DB02

Claims (11)

[Claims] 1. A color extracting apparatus for extracting a predetermined color from a color image, wherein the first numerical data is stored, and first element data constituting image information of the color image is stored in accordance with the first numerical data. A first conversion table to be converted and a second conversion table
And a second conversion table for converting second element data constituting image information of the color image in accordance with the second numerical data, and a third numerical data for storing third numerical data. A third conversion table for converting the third element data constituting the image information of the color image according to the above, and three types of data respectively converted by the first, second and third conversion tables. A color extraction device comprising: a synthesis processing unit that forms a synthesized image. 2. The color extracting apparatus according to claim 1, wherein said first, second and third numerical data correspond to colors to be extracted. 3. A library unit storing the first, second, and third numerical data as one set, and storing a plurality of sets of the one set of data for a plurality of different colors, and the plurality of sets stored in the library unit. And selecting means for selecting one set from the set and storing the set of first, second, and third numerical data in the first, second, and third conversion tables, respectively. Item 3. The color extraction device according to Item 2. 4. A display section for displaying a color image, a sampling area specifying section for specifying a specific area of the color image displayed on the display section as a sampling area, and a color model conversion section for converting the color image of the sampling area into a color model. A histogram creation unit for creating a histogram for each component of the color image information obtained by the conversion and displaying the histogram on the display unit; and displaying the geometrical figure based on the first, second, and third numerical data on the display unit. And a manipulation unit for transforming the geometrical figure displayed on the display unit. The first and the second are based on the geometrical figure after the transformation.
2. The color extracting device according to claim 1, wherein the second and third numerical data are changed. 5. The color extracting apparatus according to claim 4, wherein said display unit displays the histogram and the geometric figure in a superimposed manner. 6. A color extraction method for extracting a predetermined color from a color image, the method comprising: converting first element data constituting image information of the color image according to first numerical data.
And a second step of converting the second element data constituting the image information of the color image according to the second numerical data.
And a third step of converting third element data constituting image information of the color image according to third numerical data.
And a synthesizing step of synthesizing the data converted in the first, second and third conversion steps to form a synthesized image. 7. A color extracting method according to claim 6, wherein said first, second and third numerical data correspond to colors to be extracted. 8. A display step of displaying a color image, a sampling area designating step of designating a specific area of the displayed color image as a sampling area, and converting the color image of the sampling area by a color model conversion unit. A histogram creation step of creating a histogram for each component of the obtained color image information and displaying the histogram on a display unit;
A conversion table display processing step of displaying a geometric figure based on the first, second, and third numerical data on the display unit;
A numerical data changing step of deforming the geometric figure displayed on the display unit with the operation unit and changing the first, second, and third numerical data based on the deformed geometric figure. The color extraction method according to claim 6, wherein 9. The color extraction method according to claim 8, wherein the display unit displays the histogram and the geometric figure in a superimposed manner. 10. A storage medium storing a color extraction processing program for extracting a predetermined color from a color image, the storage medium storing a color extraction processing program for converting first element data constituting image information of the color image in accordance with first numerical data. 1 conversion processing, a second conversion processing for converting the second element data forming the image information of the color image according to the second numerical data, and forming the image information of the color image according to the third numerical data. A processing program including a third conversion process for converting the third element data and a synthesis process for forming a synthesized image by synthesizing the data converted by the first, second, and third conversion processes is stored. A storage medium characterized by the above-mentioned. 11. A display process for displaying a color image, a sampling region designating process for designating a specific region of the displayed color image as a sampling region, and converting a color image of the sampling region by a color model converting unit. A histogram is created for each component of the obtained color image information and displayed on a display unit, and a geometric figure based on the first, second, and third numerical data is displayed on the display unit. Conversion table display processing,
A processing program including a changing process for changing the first, second, and third numerical data based on the deformed geometrical figure by storing the geometrical figure displayed on the display unit using the operating unit is stored. The storage medium according to claim 10, wherein: