JPH0383476A - Specified color extracting method - Google Patents

Abstract

PURPOSE:To accurately identify the fine distinguishment of colors by multiplying a coefficient to be calculated according to difference respectively between the three primary color components of a reference color and the three primary color components of a specified color to be extracted, and afterwards, adding respective them for each primary color. CONSTITUTION:The coefficient calculated according to the difference respectively between the three primary color components of the reference color and the three primary color components of the specified color to be extracted is multiplied for each primary color and afterwards, added to three primary color signals obtained by image picking-up an object X by a color television camera 1. This, sum of products operation is executed concerning three specified colors 1-3 and with an obtained sum of products operation value as a pair, the specified colors 1-3 are extracted. Thus, when the identification of the colors is executed on a space where the obtained difference between the reference color and the specified color is satisfactorily expressed, accuracy is improved for identifying the color and the fine color difference can be also discriminated as well.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field 1] The present invention relates to a specific color extraction method for extracting a portion of an imaged object having a specific color from a color image.

[Prior art 1] Recently, in order to have a robot recognize the predetermined position of an article using a visual sensor, or to automatically inspect the appearance of a product, an image of the object to be recognized is captured using a color television camera. Efforts have been made to extract specific color parts of objects.

One such method for extracting a specific color is to set a range of the specific color to be extracted for each primary color coordinate axis in the three primary color vector space, and then determine the specific color portion based on whether the measured value falls within this rough range. There is a way to extract it.

However, with this method, there is a large difference in brightness i5'? As a result, the accuracy with which colors can be distinguished is low, and only objects with distinct color differences can be created as N-elements.

The following method is available as a method that is not affected by such brightness fluctuations. This method calculates the ratio of the three primary color components that make up the specific color you want to extract in the three primary color vector space, and then determines whether the measured value is in a region with the same ratio or similar ratio as the specific color. This is to extract the color part.

[Problem to be solved by invention F] However, in the above method, for example, even if the luminance of the three primary color components of the measured value is uniformly high, if the ratio is the same, it will be extracted as a specific color. There was a problem that it was not possible to distinguish between

The present invention has been made in view of the above-mentioned points, and an object thereof is to provide a specific color extraction method that allows even subtle color distinctions to be made with good quality.

[Means for Solving the Problem] In order to achieve the above object, the present invention extracts the three primary color signals of the reference color and the three primary color signals obtained by imaging the object to be recognized with a color television camera/hand. Multiply the coefficients found from the differences between the three primary color components of the desired specific color for each primary color, add them together, perform this product-sum calculation for at least three specific colors, and combine the obtained product-sum calculation values. This is how a specific color is extracted.

In addition, the degree of correlation with the specific color of the set of product-sum operations obtained by performing the above-mentioned product-sum operation on colors whose relationship with the specific color is known in advance is obtained by imaging the object to be recognized with a color television camera. A specific color may be extracted by converting a set of product-sum calculation values of three primary color signals into correlation data with a specific color.

[Function 1] As described above, the present invention adds three primary color components of a reference color and three primary color components of a specific color to be extracted to three primary color signals obtained by imaging an incoming object with a color television camera. By multiplying and adding the coefficients found from the differences for each primary color, the three primary color signals are mapped to a vector space consisting of the vector of the difference between the reference vector and the specific color vector in the three primary color space, and this product-sum operation is performed at least By doing this for three or more specific colors, any point in the three primary color vector space is uniquely represented by a set of product-sum calculation values, and in the space that maximizes the color difference between this reference color and the specific color. It is designed to enable color identification and to distinguish even subtle differences in color.

In addition, the degree of correlation with the specific color of the set of product-sum calculations obtained by performing the above-mentioned product-sum calculation on colors whose relationship with the specific color is known in advance is obtained by imaging the recognized N* objects with a color television camera. By converting a set of product-sum calculation values of the three primary color signals into correlation data with a specific color, it becomes possible to express how close the captured color is to the specific color, and even more accurate identification is possible.

[Example 1 An embodiment of the present invention is shown in FIGS. 1 to 4.

In this embodiment, the specific color extraction method of the present invention is applied to a device that displays on a display an image obtained by extracting a specific color portion of an object to be recognized. In addition, in this embodiment, as shown in FIG. 1, 1 to
A specific color portion is extracted for the object X in which . Note that the surface of the object X has a −-like color, and ~ In addition, in this device, after extracting the specific color part 3 is binarized using a correlation degree threshold, and 3 is individually displayed on the displays 17 to 19 for each specific color ff1i.

In the present invention, the object X was imaged by the color television camera 1. The three primary color signals are multiplied by coefficients determined from the differences between the three primary color components of the reference color and the three primary color components of the specific color to be extracted (any of specific colors 1 to 3) for each primary color, and then added. A product-sum calculation is performed for three specific colors 1 to 3, and the obtained product-sum calculation values are used as a set to extract specific colors 1 to 3. In other words, the coefficients obtained from the differences between the three primary color components of the reference color and the 3FA chrominance of the specific color to be extracted are applied to the three primary color signals obtained by imaging the object X with the color television camera 1 as described above. Performing a product-sum operation in which each primary color is multiplied and then added is a mapping of the three primary color signals to a vector space consisting of a vector of the difference between the reference color vector and the specific color vector in the three primary color vector space. The space consisting of the difference vectors attenuates the color difference between the reference color and the specific color to the maximum extent possible. This will be explained based on FIG. 2 on a two-dimensional vector space for simplicity. In FIG. 2, vector EI of the difference between reference color vector A and specific color vector B: On the other hand, if there is input vector C, performing the above-mentioned product-sum operation on input vector C means It represents the projection onto the difference vector E, and in the figure F
It is expressed as In addition, by performing a product-sum operation in which each primary color is multiplied by a coefficient determined from the difference between the three primary color components of the reference color and the three primary color components of the specific color to be extracted, and then added, it is not affected by brightness. . Performing this product-sum calculation for the three specific colors 1 to 3 means that any point in the three primary color vector space is uniquely represented by a set of product-sum calculation values. In other words, if colors are identified in a space that well represents the color difference between the reference color and the specific color obtained in this way, the color identification accuracy will be high and even subtle color differences will be distinguishable. It becomes possible. Note that here, the recognition target object By converting the set of product-sum calculation values of the three primary color signals obtained by imaging with the TV camera 1 into correlation data with a specific color, it becomes possible to express how close the captured color is to the specific color, and further improves accuracy. A high degree of discrimination is possible.

The apparatus shown in FIG. 1 will be specifically explained below. First, a color television camera 1 captures an image of the object to be recognized
convert it into a digital signal.

From this conversion data, a specific color portion is extracted using the above-described specific color extraction method of the present invention. At this time, the surface color, which is the ground color of object X, is used as the reference color, and its three primary color components are (Rb
t G bt B b), and the specific color S ×
1 to , respectively, (Rp (y G 11
1 HBl)I)% (R9isGl]z+B11z
)・(R1131G l13* B +1s) In this case, the coefficients to be multiplied by the three primary color signals obtained by imaging the object X with the color television camera "71" are determined as follows. The coefficient for R is 'I?'2t
'3, then a,=Rp,-Rb cr2=Rp2-Rb a,=R old-Rb. Similarly, the respective coefficients β1 . β2. β, and Vγ4. γ2. When determining γ, β+ = G l) + G b γ, = Bp
, -Bbβ2=Gp2-Gb γ2=Bp2-B
hβ3=Gp3-Gb γ,=Bp3-Bb. Each coefficient obtained in advance in this way is
Multipliers 41-43151-5. ．． 6. ~6. The three primary color signals obtained by imaging the object X with the color television camera 1 are multiplied by these multipliers 4. 〜43゜51〜53t61〜
6. The outputs of are added by adders 7 to 9, respectively. Here, if the three primary color components of the imaging data obtained by imaging the object X with the color television camera 1 are (Rt G t B ), then
Outputs O1-0 of each adder 7-9. is 01=fflR+β, G+γ1B 02=α2R+β2G+72B 03=Q 3R+β, G+7,B. The outputs O1-0:l of the adders 7-9 are converted into data showing the correlation with the specific colors 1-3 in a look 77 table (hereinafter referred to as L, U, T) 10.

For simplicity, an example of how to determine the value representing the degree of correlation in U and T10 will be described in a two-dimensional case with reference to FIG. In FIG. 3, the horizontal axis stones indicate the direction of the vector of the difference between the reference color and the specific color 1, and the vertical pongee 12 indicates the direction of the vector of the difference between the reference color and the specific color 2. Then, perform the above-mentioned sum-of-products calculation for colors whose relationship with the specific color is known in advance, and plot the results of the sum-of-products calculation on the plane determined by the axes e, e2, to calculate the degree of correlation for each specific color 1 and 2. 1l
Determine the in range. Note that areas a1 to a shown in FIG.
, , l+, -b have correlations of 100%, 90%, and 7, respectively.
0% and 50% areas are shown. Then, to make subsequent processing easier, assign values as shown in the table below.

Note that if it is outside the area, it is set to O. For example, if the set of measured product-sum calculation values becomes the point plotted with an x mark in FIG. 3, the output values of L, U, and T 10 will be 7.

Product-sum operation W, set of values (0+, 0z-Ot) as Ll, U, T
After inputting the data to the IO and converting it into correlation data with specific colors 1 to 3 as described above, the threshold value indicated by the correlation degree for each specific color 1 to 3 stored in the memories 14 to 16 is calculated. The above conversion output is compared with the comparators 11 to 13, and each specific color 1 to 3 is
For each case, a portion having a correlation value greater than a certain correlation degree is determined, and this data is converted into an analog signal by D/A converters 17-19 and output to each display 20-22. Therefore, each display 20-22 displays its specific color 1.
Images of the portion x and ~x3 extracted from ~3 can be displayed, respectively. Note that the images shown next to each of the displays 20 to 22 in FIG. 1 represent the images.

In addition, in the above explanation, the final processing result is only displayed on the display, but it is possible to connect an existing image processing device instead of the display and automatically measure the loss, area, etc. of a specific color y; In this case, it becomes an automatic visual inspection system that automatically performs visual inspection. Also,
In this embodiment, the case where there are three specific colors has been described, but even if there are three or more specific colors, the specific colors can be extracted with high accuracy in the same manner.

Effects of the Invention 1 As described above, the present invention provides three primary color signals obtained by capturing an image of an object with a color TV camera, the three primary colors of the reference color and the three primary color components of the specific color to be extracted. The three primary color signals are mapped to a vector space consisting of the vector of the difference between the reference vector and the specific color vector in the three primary color space by multiplying each primary color by a coefficient found from the difference between the By performing the sum operation on at least three specific colors, any point in the three primary color vector space is uniquely represented by a set of product-sum calculation values, and the color difference between this reference color and the specific color is maximized. Can distinguish colors in the space represented by , and can distinguish even subtle differences in color.

In addition, the degree of correlation with the specific color of the set of product-sum calculations obtained by performing the above-mentioned product-sum calculation on colors whose relationship with the specific color is known in advance is obtained by imaging the recognized N East object with a color television camera. By converting a set of product-sum W values of the three primary color signals into correlation data with a specific color, it is possible to express how close the captured color is to the specific color, and even higher-level discrimination is possible.

[Brief explanation of drawings]

Pt%1 Figure is a block diagram of an apparatus using the specific color extraction method of the present invention, Figure 2 is an explanatory diagram of the specific color extraction method of the present invention, and Figure 3 is a diagram illustrating the specific color extraction method of the present invention.
The figure is an explanatory diagram showing an example of how to determine the correlation degree when converting the product-sum calculation value into correlation degree data with a specific color, and FIG. 4 is a process 70-chart of the same as above. 1 is a TV camera, 41-4. ．． 5. ~53.61~6
, are multipliers, 7 to 9 are adders, 10 is a lookup table, X is an object, and 2 is a specific color portion.

Claims (2)

[Claims] (1) For each primary color, coefficients determined from the differences between the three primary color components of the reference color and the three primary color components of the specific color to be extracted are added to the three primary color signals obtained by imaging the recognition target object with a color television camera. A specific color extraction method comprising multiplying and then adding, performing this product-sum calculation for at least three specific colors, and extracting a specific color by using the obtained product-sum calculation values as a set. (2) For each primary color, coefficients determined from the differences between the three primary color components of the reference color and the three primary color components of the specific color to be extracted are added to the three primary color signals obtained by imaging the recognition target object with a color television camera. A specific color of a set of product-sum operations obtained by adding after multiplication, performing this product-sum operation on at least three or more specific colors, and performing the above-mentioned product-sum operation on colors whose relationship with the specific color is known in advance. A specific color is obtained by extracting a specific color by converting a set of product-sum calculation values of three primary color signals obtained by imaging the recognition target object with a color television camera into correlation data with the specific color, based on the degree of correlation with the object. Extraction method.