KR101289473B1 - Unlevel dyeing quanity system and method using the same - Google Patents

Abstract

PURPOSE: Unlevel dyeing quantification system and unlevel dyeing method using the same are provided to quantitatively judge a degree of unlevel dyeing of a dyed part through contour pixel image analysis. CONSTITUTION: A grayscale image conversion unit (120) converts an image of a dyeing material into an image in grayscale. A count level configuration unit (130) determines a count level according to a grayscale range. A contour pixel configuration unit (150) applies a contour pixel to level lower or level upper according to the count level. An unlevel dyeing quantification unit (160) selects an image based on a predetermined tolerance value in an image in which the contour pixel level is applied. [Reference numerals] (110) Image input unit; (120) Grayscale image conversion unit; (130) Count level configuration unit; (140) RGB image conversion unit; (150) Contour pixel configuration unit; (160) Unlevel dyeing quantification unit

Description

Unlevel dyeing quanity system and method using the same

In general, the present invention is a contour pixel (contour pixel) of the degree of uneven dyeing of the non-uniformly dyed portion of the dye dyeing laboratory or in the field. The disproportionate quantitative system for quantitative determination through image analysis and the method for quantifying disproportionate using the same, in particular, various color ranges (dark, gray, purple, black, etc.) and primary colors (Yellow, Orange, Red, etc.) The present invention relates to a disproportionate quantification system that can be accurately and objectively quantified even under conditions of inhomogeneity that are difficult to distinguish from)), and to a method for quantifying disproportionate using the same.

In the textile industry, the manufacturing stream is characterized by high pollution load and energy consumption, and there is a possibility of contraction of the existing market due to environmental regulations. In particular, the dyeing process, one of the main streams in the textile industry, removes impurities It is the step of giving the best performance in the fabric manufacturing process to give the color and performance suitable for the purpose of use.

However, most dyeing companies are small and medium-sized companies that process dyeing in the form of forestry. If reinfection occurs due to dye disproportionation and color difference, it can cause economic loss such as productivity and quality deterioration, and increase of pollutant and energy consumption.

In order to drastically reduce the troubles of dyeing companies, it is necessary to select the optimal dyes and preparations for dyeing conditions and optimize the dyeing process, and to introduce IT techniques for systematic color management of products produced in the field. It is important.

Currently, the determination of disproportionate staining of the stain is made according to the subjective judgment of the field experience. Although some automatic inspection equipment is introduced, the accuracy and the method and condition of judging the degree of uneven infection are simple and cannot be applied to various colors of the dye.

As a conventional technique for quantifying the degree of disproportionate dye dyes, for example, "A study on the analysis of fabric disproportion" published in the Journal of the Korean Society of Textile Engineers (2000), Korean Patent Registration No. 10-0933976 ( Announcement: Dec. 28, 2009) “Textile Density Measurement System and Fabric Density Measurement Method Using the Same”, Korean Patent Publication No. 10-2001-0078707 (Publication Date: Aug. 21, 2001) “Texture Representation Method Using Gabor Filter in Frequency Space and Texture-based image retrieval method ”when analyzing the disproportionate of dyed fabrics, the conditions of analysis of dark color (black, purple: color difference = 3) and primary color (orange: color difference = 8) are different, There are some cases in which the output itself does not match the reality by limiting to two cases such as disproportionate and leveling.

In addition, the area calculation method using pixel variation used in the fabric density measurement system measures 256 images of the fabric and converts the measured images into grayscale images that can be expressed in 256 tone steps as 8-bit images. The image is processed according to the standard tone (H) created by resetting the tone level to 4 levels. When processed in this way, it is possible to calculate the area in two cases, such as disproportionate and leveling, but it is impossible to calculate grayscale image conversion when the two are mixed.

In addition, the texture expression method using a Gabor filter is also far from the present invention in terms of the purpose and method of quantifying the degree of inhomogeneity.

In order to extract the smooth image information about the equalized and unbalanced portions of the chlorine, a clear image segmentation must be preceded. The image segmentation divides an image into components or objects.

The range of sessionization of this division differs depending on the problem to be solved. Generally, in order to recognize a set of adjacent pixels as one connected area, it is necessary that the areas are homogeneous. The homogeneity of the region is related to the nature of each pixel (brightness, color), the nature of the local set of pixels (geometric pattern). However, the disproportionate area of the object is difficult to extract the image information because there is no clear boundary of the image, and the automatic image conversion process is more difficult. It is no exaggeration to say that the image segmentation step in the image processing determines the ultimate success or failure of image interpretation. For this reason, there have been many studies to increase the probability of smooth image segmentation, but at present, no general image segmentation method has been developed, and image segmentation using color, knowledge based image segmentation system, and image segmentation using optimization have been actively studied recently. have.

Gabor filters have been studied a lot in pattern image segmentation. Gabor filter is named after Dennis Gabor in image processing and is a linear filter used for edge-based detection. Its superiority is recognized but it is difficult to implement and takes a long time to execute. It is not used very much.

In accordance with these problems, a method that can accurately and objectively quantify the degree of disproportionation of non-uniformly dyed portions of dyeing dyes in a dyeing laboratory or in the field is generally desired.

An embodiment of the present invention is a non-uniformity quantitative system that can be determined quantitatively through the contour pixel image analysis of the degree of non-uniformly dyed portion of the dye stained in the dyeing laboratory or field and using the same It is intended to provide a method for quantitating disproportionate salt.

According to an aspect of the present invention, a disproportionate quantification system capable of quantitatively analyzing the degree of disproportion may be provided by receiving an image of the object to which a constant load is applied and analyzing the contour pixel image. .

In this case, the image of the object may be input to the image taken by photographing the camera or the scanned image using a scanner.

On the other hand, the disproportionate quantitative system comprises an image input unit for receiving the image of the blood; A grayscale image converting unit converting the image of the chloride object into a gray scale image; A count level setting unit for determining a count level according to the grayscale range; An RGB image converting unit converting the gray scale image having the count level set into an RGB image; Set a contour pixel level value for a border portion of the pattern of the RGB converted image, and level the contour pixel according to the count level determined in gray scale based on the set contour pixel level value. Select the image based on the preset tolerance value in the contour pixel setting unit and the contour pixel level value applied image and apply the histogram for the resulting image. And a disproportionate quantitative unit for calculating a pixel value using the histogram.

In this case, the count level setting unit may set a level upper when the count level is GS <127 and set a level lower when the count level is GS> = 127.

In this case, the tolerance value may be set to 10.

According to another aspect of the present invention, the step of receiving an image of the object to which the constant load is applied; Converting the image to a grayscale image; Determining whether the grayscale is greater than or equal to 127; Setting a level lower if the grayscale is greater than or equal to 127; RGB converting the image; Setting a contour pixel for the border portion of the pattern; Applying a leveler or levelupper to the contour pixel according to the count level determined in the grayscale based on a set contour pixel value; Selecting an image based on a tolerance value set in the image to which the contour pixel value is applied; Obtaining a histogram for the resultant image, and calculating a pixel value using the histogram may be provided for the method for quantitating the disproportionate using a disproportionate quantitative system.

In this case, when the grayscale is smaller than 127, the method may further include applying a level upper.

As described above, the present invention can quantitatively determine the degree of non-uniformity of the non-uniformly dyed portion of the dye dyeing laboratory or in the field through contour pixel image analysis, and in particular, various color ranges. Can be objectively quantified even under various disproportionate conditions.

1 is a diagram showing three types of examples in general when there is a disproportionate in a stained blood stain.
2 is an overall configuration diagram of a disproportionate quantification system according to an embodiment of the present invention.
Figure 3 is a flow chart for the disproportionate quantitative method using the disproportionate quantification system according to an embodiment of the present invention.
FIG. 4 is a photograph of an actual image of a blood stain and a contour pixel level value.
FIG. 5 is a diagram of quantifying disproportionated pixels by contour pixel image analysis in a form of A and B disproportionate forms (wrinkles, lines, and dark spots).

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, configurations and operations of embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram showing three types of examples in general when there is a disproportionate in a stained blood stain. Type 1 is when the specific part is dyed dark and the remainder is uniform. Type 2 is when the specific part is not dyed lightly or not at all, and the remaining part is uniform. Type 3 is when the two above are combined. to be.

Disproportionate condition of dyed stains Type 1 and type 2 can be calculated by the conventional pixel conversion method, but type 3 is a combination of two cases, so that the uncontaminated area is converted into a grayscale image and converted into pixels. It is impossible to calculate.

For this reason, edge-based image segmentation and region-based image segmentation cannot be used separately in the type 3 area calculation method using pixel variation.

The edge-based image segmentation method is not easy to find the boundary between different patterns because the image does not have much difference in brightness value, so it cannot be used as a result of image segmentation alone, and is based on region-based image segmentation. Since the method compares similarities between regions and merges similar regions to generate an image, image segmentation should be applied by applying a combination of region-based and edge chains detected and detected as edge-based.

Therefore, a new image analysis method that combines the two methods must be used for accurate analysis.

In addition, a pre-filtering process may be necessary to better express the image transformation.

To do this, it is necessary to determine the count level according to the grayscale range in the inputted chlorine image. This means that in the case of unevenly stained stains, some of the stains are dark or some of them are not stained at all. In addition, since the above two cases are mostly combined, heterogeneity may vary according to the image selection site. Therefore, it is necessary to restore the grayscale image back to the RGB image to set the contour pixel level value for the border portion of the pattern.

Based on the set contour pixel level value, the contour pixel is lowered or leveled up according to the count level determined by the gray scale. Can be applied as

To this end, the disproportionate quantitative system according to an embodiment of the present invention obtains an image of the disproportionate contaminants through scanning or image capturing, and accurately corrects various disproportionate forms through contour pixel image analysis. And quantitative analysis.

More specifically, the heterogeneous quantitative quantitative system according to an embodiment of the present invention uses a contour pixel image analysis method to determine the degree of nonuniformity of a non-uniformly dyed portion of a chlorine in various color ranges (dark color (Navy). , Grey, Purple, Black, etc.) and primary conditions (Yellow, Orange, Red, etc.) and disproportionate conditions (Type 1: darkened specific areas and uniform remainder, Type 2: light or dark areas) If it is not stained and the remainder is uniform, it can be accurately and objectively quantified even under type 3: where the above two are combined.

2 is an overall configuration diagram of a disproportionate quantification system according to an embodiment of the present invention.

2, the disproportionate quantification system according to an embodiment of the present invention is an image input unit 110, grayscale image conversion unit 120, count level setting unit 130, RGB image conversion unit 140, It may be composed of a contour pixel setting unit 150 and an uneven salt measuring unit 160.

The image input unit 110 receives a photographed image of the blood stain. The image input unit 110 may take an image of the object to be received by using a camera or the like and receive it. The image input unit 110 may receive a resolution, an applied load, and the like, and set it.

Although not shown in the drawing, the image input unit 110 may be installed as some components of the dyeing machine and apply a predetermined load to the sample of the dyeing completed, and then receive an image through photographing. In addition, the image input unit 110 may be installed separately from the dyeing machine to apply a predetermined load to the dye is completed, and then take an image to receive the photographed image.

In the disproportion quantitative system according to an exemplary embodiment of the present invention, the image input unit 110 receives a photograph by capturing a photograph of the photograph of the object, but the object may be input as an image file by scanning the object.

The grayscale image converter 120 may convert the image of the blood stain into a gray scale image.

The count level setting unit 130 determines a count level according to the grayscale range. If the gray scale range is less than 127 (GS <127), most of the image color is in the dark range, and if the gray scale range is 127 or more (GS> = 127), most of the image color is in the light range.

The count level setting unit 130 may include a bright range in the image conversion by setting a level upper when the count level is set to a dark range such as GS <127, and the count level is GS> = 127. For example, if you have a lot of bright scale ranges set, you can set the level lower to include dark ranges in the image transition.

The level setting of the count level setting unit 130 can be selected through the level upper (level upper), even if the specific portion of the disproportion of the effluent is light or not dyed at all through the level setting of the counting level setting unit 130 In this case, it is possible to select through the level lower setting, so that the disproportionate of the two types of contaminants can be quantified.

The RGB image converter 140 may convert a gray scale image having a count level set into an RGB image.

The contour pixel setting unit 150 may set a contour pixel level value for a boundary portion of the pattern of the image converted into RGB.

The contour pixel setting unit 150 raises the contour pixel according to the count level determined by the grayscale based on the set contour pixel level value. It can be applied as lower or level upper.

The disproportionate quantitator 160 selects an image based on a preset tolerance value in the image to which the contour pixel level is applied, obtains a histogram of the resultant image, and the histogram. The pixel value can be calculated using. In this case, the tolerance value may apply to 10.

Next, the sterility quantification method using the sterility quantification system according to an embodiment of the present invention will be described.

Figure 3 is a flow chart for the disproportionate quantitative method using the disproportionate quantification system according to an embodiment of the present invention.

Referring to FIG. 3, first, a constant load is applied to a constant area of the object to receive an image of the object of the object (S110). When a certain load is applied, the surface texture of the blood and the image of the blood can be uniformly input.

Next, the image of the photographed chlorine is converted to a gray scale image (S120).

Thereafter, it is determined whether the count level according to the gray scale range is greater than or equal to 127 (S130). Divide the range from black to white by 0 to 255 and set the 127 range, which is half of it, as the reference for the count level. If the gray scale range is less than 127 (GS <127), most of the image color is in the dark range, and if the gray scale range is 127 or more (GS> = 127), most of the image color is in the light range.

If the count level is set to a dark range such as GS <127, the level upper is set to include a bright range in the image transition (S140), and the count range has a lot of bright scale ranges such as GS> = 127. If it is set, a dark range may also be included in the image conversion by setting a level lower (S150).

The sterility quantification method using the sterility quantification system according to an embodiment of the present invention is selected through the level upper (level upper) setting even when a specific portion of the sterile salt of the phosphate is lightly stained or not stained at all through the level setting. This is possible, and even if the dye is darkly stained among the disproportion of the chlorine can be selected through the level lower (level lower) setting can also quantify the disproportion of the complex of the two complex forms.

Next, the grayscale image is converted back into an RGB image (S160).

Next, a contour pixel level value for a border portion of the pattern is set (S170).

Thereafter, the contour pixel is level lowered or leveled up according to the count level determined in grayscale based on the set contour pixel level value. level upper) (S180).

Next, the image is selected based on a preset tolerance in the image to which the contour pixel level value is applied (S190).

Finally, a histogram of the resulting image is obtained and pixel values are calculated using the histogram (S200).

FIG. 4 is a photograph of an actual image of a blood stain and a contour pixel level value. If you apply contour pixel level values to uneven areas that are difficult to distinguish on the actual image, the blue-green series is a balanced region and the red-white series It can be seen that it appears as an uneven portion. Of course, the color sequence can be set according to the contour pixel level value.

FIG. 5 is a diagram of quantifying disproportionate pixels by contour pixel image analysis in a form of A and B disproportionate forms (wrinkles and lines (type 3) and dark spots gathered in one form (type 2)). . The disproportionate pixel of A is 19,000, and the disproportionate pixel of B is 32300, which shows that the disproportionate of B-chloride is more severe. Although it is difficult to compare the degree of disproportion due to the different forms of disproportion, it is possible to clearly distinguish the degree of disproportion when quantitatively obtaining disproportion pixels through the analysis of the contour pixel image. The sample was 20 g in weight and 20 × 20 cm in size, and was stained at 60 degrees with 30 g / l of forget-me-not (Na ₂ SO ₄ ) and 20 g / l of soda (Na ₂ CO ₃ ). The applied load is 50kg / m ² and the resolution is 200dpi.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, It will be understood. The embodiments described above are therefore to be considered in all respects as illustrative and not restrictive.

100: disproportionate determination system
110: image input unit
120: grayscale image conversion unit
130: count level setting unit
140: RGB image conversion unit
150: contour pixel setting unit
160: disproportionate determination unit

Claims (8)

As a disproportionate quantification system that can quantitatively analyze the form of disproportionate,
An image input unit configured to receive an image of a blood stain;
A grayscale image converting unit converting the image of the chloride object into a gray scale image;
A count level setting unit for determining a count level according to the grayscale range;
An RGB image converting unit converting the gray scale image having the count level set into an RGB image;
Set a contour pixel level value for a border portion of the pattern of the RGB converted image, and level the contour pixel according to the count level determined in gray scale based on the set contour pixel level value. Or the contour pixel setting section to be applied as level upper
A non-uniformity measurement unit that selects an image based on a preset tolerance value in the image to which the contour pixel level value is applied, obtains a histogram of the resultant image, and calculates a pixel value using the histogram. Aseptic salt quantification system comprising a. The method of claim 1,
The image of the chloride,
Disinfectant quantification system characterized in that the camera receives an image or receives an image scanned using a scanner.
delete The method of claim 1,
The count level setting unit,
If the count level is GS <127, set a level upper,
When the count level is GS> = 127, a level lower level is set.
The method of claim 1,
The tolerance value quantitative system, characterized in that 10.
As a quantitative determination method of sterile salt using a sterile quantitative determination system,
Receiving an image of the object to which the constant load is applied;
Converting the image to a grayscale image;
Determining whether a count level according to the grayscale range is greater than or equal to 127;
Setting a level lower if the count level is greater than or equal to 127;
RGB converting the image;
Setting a contour pixel for the border portion of the pattern;
Applying a leveler or levelupper to the contour pixel according to the count level determined in the grayscale based on a set contour pixel value;
Selecting an image based on a tolerance value set in the image to which the contour pixel value is applied;
Obtaining a histogram for the resultant image, and calculating the pixel value using the histogram, a disproportion quantitative method using a disproportionate quantification system.
The method according to claim 6,
If the count level is less than 127, aseptic salt quantification method using a sterile salt quantitative system further comprising the step of applying a level upper.
The method according to claim 6,
The image may include:
Disinfectant quantification method using a disinfectant quantification system that receives an image by taking a camera or receives an image scanned using a scanner.

Images