JP2005130000A - Image processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing apparatus capable of reliably extracting a target color without being affected by illumination conditions or color change of an object.
An RGB data storage unit that stores captured color image data as RGB (R: red, G: green, B: blue) data, and RGB data stored in the RGB data storage unit are converted into HSV (H : Hue, S: Saturation, V: Luminance) RGB / HSV conversion means for conversion to data, HSV data storage means for storing HSV data converted by the RGB / HSV conversion means, and HSV data storage means A gray-scale image extracting unit that compares the stored HSV data with preset extraction color reference data and obtains gray-scale image data by calculation.
[Selection] Figure 2

Description

  The present invention relates to an image processing apparatus that extracts an arbitrary color from captured color image data. In particular, RGB (R: red, G: green, B: blue) data is converted into HSV (H: hue, S: chromatic). The present invention relates to a device that is devised so that desired grayscale image data can be obtained instead of binarized image data by performing predetermined calculation processing after conversion into data.

A conventional image processing apparatus has the following general configuration. First, a color image taken by an imaging means such as a color camera is stored as RGB (R: red, G: green, B: blue) data in a memory area in the computer. A predetermined color extraction process is performed from the stored RGB (R: red, G: green, B: blue) data.

That is, for each of the three primary colors R: red, G: green, and B: blue, the data range of the color to be extracted is set, and RGB (R: red, G: stored in the memory area in the computer). It is determined whether or not the data is within the data range of the set extraction color for the data (green, B: blue). This will be described with reference to FIG. FIG. 6A shows an example when the extraction upper limit value is larger than the extraction lower limit value, and FIG. 6B shows an example when the extraction upper limit value is smaller than the extraction lower limit value.

  First, in the case shown in FIG. 6A, the two extraction conditions of “the color level value of the measured value is equal to or higher than the extraction lower limit value” and “the color level value of the measured value is equal to or lower than the upper limit value of extraction” are satisfied. It is determined whether it is within the range, “1”, otherwise “0”. Similarly, in the case shown in FIG. 6B, either one of “the color level value of the measured value is not more than the extraction upper limit value” and “the color level value of the measured value is not less than the extraction lower limit value” is satisfied. It is determined whether or not to be performed, and is “1” if it is within the range, and “0” if it is outside the range.

In any case, the determination result is output as a binary value, that is, “1” or “0”. Further, the same determination is performed for all the pixels to obtain binarized data. Then, using the obtained binarized data, a binarized image is created and used as an extracted image.

Specifically, for example, there is a color image as shown in FIG. In this case, for example, four color fluorescent pens are photographed by a color camera. Each color highlighter includes a main body 101 and a cap 103, and a clip 105 is attached to the cap 103. Part of the cap 103, specifically, the upper end head portion, the lower end portion, and the hook 105 portion are colored, and the other portions are black.
In the drawing, for convenience, all are displayed in “white / black”, but in actuality, the portion indicated by the symbol “O” is “orange”, the portion indicated by the symbol “P” is “pink”, and the symbol “ The portion indicated by “G” is “green”, the portion indicated by the symbol “Y” is “yellow”, and the other background color (in this case, black) appears as it is.
Then, for example, by performing image processing for extracting “orange” on the color image shown in FIG. 7, binary image data as shown in FIG. 8 can be obtained. In the binarized image shown in FIG. 8, only the portion that has been binarized and determined to be “1” is displayed in “white”, and the other portions are displayed in “black”. In FIG. 7, the orange portion indicated by the symbol “O” is displayed as “white” in FIG. 8, and at the same time, some confusing portions are also displayed as “white”.

Although it seems that there is no direct relationship with the present invention, there are, for example, Patent Document 1, Patent Document 2, Patent Document 3, Patent Document 4 and the like that disclose this type of image processing apparatus.

JP 2002-298083 A JP 2002-163650 A JP 2002-131133 A JP 2002-49916 A

The conventional configuration has the following problems.
That is, in the case of a conventional image processing apparatus, whether or not a data range of colors to be extracted is set for the three primary colors RGB (R: red, G: green, B: blue) and falls within the data range. I am trying to determine. However, in this case, the target color may not be extracted due to lighting conditions or color change of the object. This is because even if the object has the same color, the luminance and saturation change depending on the lighting conditions.
For such a problem, when extracting a color, it is necessary to perform a separate process so that the degree of extraction with respect to luminance and saturation can be adjusted, which is a very complicated process.
In addition, since the extracted image data is a binary image, only the two colors of the portion displayed in “white” and the portion displayed in “black” are expressed. There was a problem that it was extremely difficult to recognize things.
Furthermore, if the object cannot be accurately identified, colors other than the object may be determined as the extracted color range, and as a result, colors other than the target color will be extracted. .

  The present invention has been made on the basis of such points, and the object of the present invention is to reliably extract the target color without being affected by the lighting conditions or the color change of the object. Another object is to provide an image processing apparatus.

In order to achieve the above object, an image processing apparatus according to claim 1 of the present invention comprises RGB data storage means for storing captured color image data as RGB (R: red, G: green, B: blue) data; RGB / HSV conversion means for converting RGB data (R: red, G: green, B: blue) stored in the RGB data storage means into HSV (H: hue, S: saturation, V: luminance) data; HSV data storage means for storing HSV (H: hue, S: saturation, V: luminance) data converted by the RGB / HSV conversion means; and HSV (H: hue) stored in the HSV data storage means. , S: chroma, V: luminance) data and grayscale image data extraction means for comparing grayscale data and preset grayscale data to obtain grayscale image data. It is.
An image processing apparatus according to claim 2 is the image processing apparatus according to claim 1, wherein the grayscale image data extracting means includes the HSV (H: hue, S: saturation, V: luminance) data and an extracted color reference. Difference data calculation means for calculating a difference (ΔH, ΔS, ΔV) from the data, and difference data (ΔH, ΔS, ΔV) calculated by the difference data calculation means are subjected to aggregation calculation processing to obtain grayscale image data. And an aggregating operation processing means.
According to a third aspect of the present invention, in the image processing apparatus according to the second aspect, the grayscale image data extracting means adds an arbitrary coefficient to the difference data (ΔH, ΔS, ΔV) obtained by the difference data calculating means. Is multiplied by the coefficient multiplication means for adjusting the extraction sensitivity, and the aggregation calculation processing means is an aggregation calculation based on the coefficient multiplication difference data (ΔH ′, ΔS ′, ΔV ′) obtained by the coefficient multiplication means. Processing is performed to obtain grayscale image data.
According to a fourth aspect of the present invention, in the image processing apparatus according to the third aspect, the coefficient is an arbitrary value equal to or greater than “0”.
An image processing apparatus according to claim 5 is the image processing apparatus according to claim 3 or 4, wherein the aggregation calculation processing means performs an aggregation calculation process based on the following expression to obtain a gray value. It is characterized by being.
(N−ΔH ′) × (N−ΔS ′) × (N−ΔV ′) / N ²
N: Number of gradations

As described above in detail, the image processing apparatus according to the present invention can provide the following effects.
First, not based on a binarized image based on RGB (R: red, G: green, B: blue) data as in the past, but based on HSV (H: hue, S: saturation, V: luminance) data. As a result, it is possible to clearly identify the object without being affected by lighting conditions or color changes of the object, and approximate the extracted color to other colors. Can be grasped visually.
Further, the grayscale image extracting means includes difference data calculating means for calculating differences (ΔH, ΔS, ΔV) between the HSV data and the extracted color reference data, and difference data (ΔH, ΔS, ΔV) is composed of aggregation calculation processing means for obtaining grayscale image data by performing aggregation calculation processing, and the difference data (ΔH, ΔS, ΔV) obtained by the difference data calculation means is multiplied by an arbitrary coefficient. When coefficient multiplication means for adjusting the extraction sensitivity is provided, it is possible to adjust the extraction sensitivity as appropriate to obtain clearer gray image data.
In addition, it is possible to set and adjust the extraction color by appropriately adjusting various numerical values, for example, coefficients, etc. while viewing the grayscale image data, so that the color to be extracted can be easily set and obtained. .
For example, when such an image processing apparatus is applied to various inspections, for example, inspections for determining whether or not various electrical components are correctly mounted on a substrate, a highly accurate inspection is performed. It will be easy to do.

Hereinafter, an embodiment of the present invention will be described with reference to FIGS.
FIG. 1 is a diagram conceptually showing the configuration of an image processing system including an image processing apparatus according to an embodiment of the present invention. First, there is a color camera 1 as an imaging means. A color image is obtained by photographing an object with this color camera. The color camera 1 is connected to a CPU (Central Processing Unit) 5 through an image input board 3.

  The CPU 5 is connected to a hard disk drive (HDD) 7, a random access memory (RAM) 9, and a flat panel display (FPD) 11. Further, color extraction processing software as an entity of the image processing apparatus according to the present embodiment is stored in the HDD 7. At the time of processing, it is loaded from the HDD 7 onto the RAM 9 according to a command from the CPU 5 and executed.

  A color image taken by the color camera 1 is stored in the color memory space on the RAM 9 via the image input board 3. Then, predetermined image processing is performed by the CPU 5, and desired grayscale extracted image data is obtained. Further, the obtained grayscale extracted image data is displayed on the FPD 11.

Next, the substance of the color extraction processing software stored in the HDD 7 will be described with reference to FIG.
First, there is a color memory R (red) channel 21, a color memory G (green) channel 23, and a color memory B (blue) channel 25. These color memory R (red) channel 21, color memory G (green) channel 23, In the color memory B (blue) channel 25, color image data photographed by the color camera 1 is stored as RGB (R: red, G: green, B: blue) data, respectively.

In addition, color space conversion processing means 27 is provided, and RGB (R: red, The G: green, B: blue) data is converted into HSV (H: hue, S: saturation, V: luminance) data by the color space conversion processing means 27. The converted HSV (H: Hue, S: Saturation, V: Luminance) data is stored in a color memory H (hue) channel 29, a color memory S (saturation) channel 31, and a color memory V (luminance) channel 33. Will be stored respectively.

Further, there is a difference calculation processing means 35. The difference calculation processing means 35 includes the extracted color reference data 37 set and inputted in advance, the color memory H (hue) channel 29, and the color memory S (saturation) channel. 31 and HSV (H: Hue, S: Saturation, V: Luminance) data respectively stored in the color memory V (luminance) channel 33 are compared with the following equations (I), (II), ( Difference data (ΔH, ΔS, ΔV) as shown in III) is calculated. At this time, the color data value of the extracted color may be larger than the color data value of the image. In this case, the value is “−”, and thus the difference value is an absolute value.
Hue difference value ΔH = | Hue value of image−Hue value of extracted color | ---- (I)
Saturation difference value ΔS = | Saturation value of image−Saturation value of extracted color | —―― (II)
Brightness difference value ΔV = | Brightness value of image−Brightness value of extracted color | --- (III)
The difference data (ΔH, ΔS, ΔV) is shown in FIG. 3A.

Further, there is a sensitivity calculation processing means 39, and this sensitivity calculation processing means 39 is preset and inputted to the difference data (ΔH, ΔS, ΔV) calculated and calculated by the difference calculation processing means 35. By multiplying the sensitivity data 41, coefficient multiplication difference data (ΔH ′, ΔS ′, ΔV ′) as shown in the following equations (IV), (V), (VI) is calculated. Further, the sensitivity coefficient is a value equal to or greater than “0”. For example, 0 indicates a difference value of 0 and 10 indicates a difference value of 10 times.
Coefficient multiplication hue difference value ΔH ′ = hue difference value ΔH × hue sensitivity coefficient (0 or more) --- (IV)
Coefficient Multiply Saturation Difference Value ΔS ′ = Saturation Difference Value ΔS × Saturation Sensitivity Coefficient (0 or More) --- (V)
Coefficient multiplication luminance difference value ΔV ′ = luminance difference value ΔV × luminance sensitivity coefficient (0 or more) --- (VI)

In addition, there is a color component aggregation calculation processing unit 43, and the color component aggregation calculation processing unit 43 is for the coefficient multiplication difference data (ΔH ′, ΔS ′, ΔV ′) calculated by the sensitivity calculation processing unit 39. The processing shown in the following formula (VII) is performed, thereby calculating grayscale data and obtaining desired grayscale image data 45 based on the grayscale data.
Gray value = (N−ΔH ′) × (N−ΔS ′) × (N−ΔV ′) / N2
――― (VII)
However,
N: number of gradations, 255 in this embodiment
FIG. 3B shows the relationship between the gray value calculated by the above formula (VII) and the measured value.

Based on the above configuration, the operation will be described with reference to the flowchart of FIG.
First, in step S1, a color image color space conversion process is executed. As a result, RGB (R: red, G: green, B: blue) data is converted to HSV (H: hue, S: saturation, V: luminance) data. Hereinafter, based on the converted HSV (H: Hue, S: Saturation, V: Luminance) data, predetermined image processing is executed for all pixels in the X and Y axis directions. .

That is, the process proceeds to step S2, and it is determined whether or not image processing has been completed for all of the Y-axis addresses. If the predetermined image processing has been completed for all the pixels, the process proceeds to step S3, the creation of the extracted image is executed, and the process ends thereafter. On the other hand, if it is determined that the process has not been completed, the process proceeds to step S4. In step S4, it is determined whether or not image processing has been completed for the X-axis address. If completed, the process proceeds to step S5, Y-axis address increment processing is executed, and then the process returns to step S1. On the other hand, if it is determined that the process has not been completed, the process proceeds to step S6.
Note that the contents of the most important image processing are after step S6.

  That is, in step S6, hue difference processing is executed. That is, the hue difference data (ΔH) is calculated. Next, the process proceeds to step S7, and hue sensitivity calculation processing is executed. That is, the hue difference data ΔH is multiplied by an arbitrary coefficient to calculate coefficient multiplied hue difference data (ΔH ′). Next, it transfers to step S8 and a saturation difference calculation process is performed. That is, the saturation difference data (ΔS) is calculated. Next, it transfers to step S9 and performs a saturation sensitivity calculation process. That is, the coefficient saturation saturation difference data (ΔS ′) is calculated by multiplying the saturation difference data (ΔS) by an arbitrary coefficient.

  Next, it transfers to step S10 and a brightness | luminance difference calculation process is performed. That is, luminance difference data (ΔV) is calculated. Next, it transfers to step S11 and performs a brightness sensitivity calculation process. That is, the luminance difference data (ΔV) is calculated by multiplying the luminance difference data (ΔV) by an arbitrary coefficient. Through the processing so far, coefficient multiplication difference data (ΔH ′, ΔS ′, ΔV ′) relating to an arbitrary pixel has been obtained.

  Next, the process proceeds to step S12, and color component aggregation calculation processing is executed. That is, a process for calculating a gray value based on the formula (VII) is executed. Next, the process proceeds to step S13, an X-axis address increment process is executed, and the process returns to step S4. Thereafter, the same processing is repeatedly executed to execute predetermined image processing on all the pixels in the X and Y axis directions, thereby calculating a gray value and obtaining a desired gray level. Value image data is obtained.

  An example of grayscale image data obtained by the above processing is shown in FIG. FIG. 5 corresponds to the color image of FIG. 7 used in the description of the conventional example. When the grayscale image data according to this embodiment shown in FIG. 5 is compared with the conventional binarized image data shown in FIG. 8, in the case of the grayscale image data according to this embodiment, the extracted color (in this case) Can be seen that the color closer to “orange” is displayed in white and the color farther away is displayed in black. Specifically, green and yellow, which are colors close to orange, are displayed whiter than pink, which is a color distant from orange.

As described above, according to this embodiment, the following effects can be obtained.
First, not based on a binarized image based on RGB (R: red, G: green, B: blue) data as in the past, but based on HSV (H: hue, S: saturation, V: luminance) data. As a result, it is possible to clearly identify the object without being affected by lighting conditions or color changes of the object, and approximate the extracted color to other colors. Can be grasped visually.
Further, the grayscale image extracting means includes difference data calculating means for calculating differences (ΔH, ΔS, ΔV) between the HSV data and the extracted color reference data, and difference data (ΔH, And an aggregation calculation processing means for obtaining grayscale image data by performing an aggregation calculation process on ΔS, ΔV), and multiplying the difference data (ΔH, ΔS, ΔV) obtained by the difference data calculation means by an arbitrary coefficient In the case where coefficient multiplication means for adjusting the extraction sensitivity is provided, it is possible to obtain clearer grayscale image data by appropriately adjusting the extraction sensitivity.
Also, by adjusting various numerical values (for example, coefficients, etc.) while looking at the grayscale image data, the extraction color can be set and adjusted, so that the color to be extracted can be easily set and obtained. .

The present invention is not limited to the one embodiment.
For example, in the above-described embodiment, the case where the sensitivity calculation process is executed has been described as an example, but an embodiment in which the sensitivity calculation is not executed is also within the scope of the present invention.
In addition to the color camera as shown, various known devices can be used as the imaging means.

  The present invention relates to an image processing apparatus that extracts an arbitrary color from captured color image data. In particular, RGB (R: red, G: green, B: blue) data is converted into HSV (H: hue, S: saturation). , V: Luminance) data is converted into data and processed so that desired grayscale image data can be obtained without being affected by lighting conditions or color change of the object. For example, it is effectively used in various inspections.

1 is a diagram illustrating an embodiment of the present invention, and is a block diagram illustrating a configuration of an image processing system including an image processing apparatus. 1 is a block diagram illustrating a configuration of an image processing apparatus according to an embodiment of the present invention. FIG. 3A is a diagram illustrating an embodiment of the present invention, FIG. 3A is a diagram for explaining the concept of color extraction, and FIG. 3B is a diagram for explaining grayscale values. 1 is a diagram illustrating an embodiment of the present invention, and is a flowchart for sequentially explaining the contents of image processing by an image processing apparatus. FIG. It is a figure which shows one embodiment of this invention, and is a figure which shows the grayscale image data obtained by the image processing by an image processing apparatus. FIGS. 6A and 6B are diagrams for explaining the concept of conventional image processing. It is a figure which shows a prior art example, and is a figure which shows the imaged color image. It is a figure which shows a prior art example, and is a figure which shows the binarized image data after image processing.

Explanation of symbols

1 Color Camera 3 Image Input Board 5 CPU
7 HDD
9 RAM
11 FPD
21 Color memory R (red) channel 23 Color memory G (green) channel 25 Color memory B (blue) channel 27 Color space conversion processing means 29 Color memory H (hue) channel 31 Color memory S (saturation) channel 33 Color memory V (luminance) channel 35 Difference calculation processing means 39 Sensitivity calculation processing means 43 Color component aggregation calculation processing means

Claims (5)

RGB data storage means for storing captured color image data as RGB (R: red, G: green, B: blue) data;
RGB / HSV conversion means for converting RGB data (R: red, G: green, B: blue) stored in the RGB data storage means into HSV (H: hue, S: saturation, V: luminance) data; ,
HSV data storage means for storing HSV (H: hue, S: saturation, V: luminance) data converted by the RGB / HSV conversion means;
The light and shade to obtain light and shade image data by comparing and calculating HSV (H: hue, S: saturation, V: luminance) data stored in the HSV data storage means and preset extraction color reference data. Image data extraction means;
An image processing apparatus comprising: The image processing apparatus according to claim 1.
The grayscale image data extraction means includes:
Difference data calculation means for calculating a difference (ΔH, ΔS, ΔV) between the HSV (H: hue, S: saturation, V: luminance) data and the extracted color reference data;
Aggregation calculation processing means for obtaining density image data by performing aggregation calculation processing on the difference data (ΔH, ΔS, ΔV) calculated by the difference data calculation means;
An image processing apparatus comprising: The image processing apparatus according to claim 2.
The grayscale image data extraction means includes:
Coefficient multiplication means for adjusting the extraction sensitivity by multiplying the difference data (ΔH, ΔS, ΔV) obtained by the difference data calculation means by an arbitrary coefficient,
An image processing apparatus characterized in that the aggregation calculation processing means obtains grayscale image data by performing aggregation calculation processing based on the coefficient multiplication difference data (ΔH ′, ΔS ′, ΔV ′) obtained by the coefficient multiplication means. The image processing apparatus according to claim 3.
The image processing apparatus, wherein the coefficient is an arbitrary value equal to or greater than “0”. The image processing apparatus according to claim 3 or 4,
The image processing apparatus characterized in that the aggregation calculation processing means performs an aggregation calculation process based on the following formula to obtain a gray value.
(N−ΔH ′) × (N−ΔS ′) × (N−ΔV ′) / N ²
N: Number of gradations