KR101757071B1 - Method of image segmentation assisted entropy - Google Patents
Method of image segmentation assisted entropy Download PDFInfo
- Publication number
- KR101757071B1 KR101757071B1 KR1020150173731A KR20150173731A KR101757071B1 KR 101757071 B1 KR101757071 B1 KR 101757071B1 KR 1020150173731 A KR1020150173731 A KR 1020150173731A KR 20150173731 A KR20150173731 A KR 20150173731A KR 101757071 B1 KR101757071 B1 KR 101757071B1
- Authority
- KR
- South Korea
- Prior art keywords
- image
- entropy
- value
- brightness
- pixel
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims description 32
- 238000003709 image segmentation Methods 0.000 title description 9
- 238000009826 distribution Methods 0.000 claims abstract description 22
- 238000001000 micrograph Methods 0.000 claims description 2
- 238000003672 processing method Methods 0.000 abstract description 2
- 239000000463 material Substances 0.000 description 10
- 238000012545 processing Methods 0.000 description 5
- BQCADISMDOOEFD-UHFFFAOYSA-N Silver Chemical compound [Ag] BQCADISMDOOEFD-UHFFFAOYSA-N 0.000 description 4
- 239000002042 Silver nanowire Substances 0.000 description 4
- 238000010191 image analysis Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000000873 masking effect Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000001878 scanning electron micrograph Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 229910021536 Zeolite Inorganic materials 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000002041 carbon nanotube Substances 0.000 description 1
- 229910021393 carbon nanotube Inorganic materials 0.000 description 1
- 239000013078 crystal Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- HNPSIPDUKPIQMN-UHFFFAOYSA-N dioxosilane;oxo(oxoalumanyloxy)alumane Chemical compound O=[Si]=O.O=[Al]O[Al]=O HNPSIPDUKPIQMN-UHFFFAOYSA-N 0.000 description 1
- 238000003618 dip coating Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000003913 materials processing Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 239000012528 membrane Substances 0.000 description 1
- 239000002808 molecular sieve Substances 0.000 description 1
- 239000002070 nanowire Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- URGAHOPLAPQHLN-UHFFFAOYSA-N sodium aluminosilicate Chemical compound [Na+].[Al+3].[O-][Si]([O-])=O.[O-][Si]([O-])=O URGAHOPLAPQHLN-UHFFFAOYSA-N 0.000 description 1
- 239000010457 zeolite Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G02—OPTICS
- G02B—OPTICAL ELEMENTS, SYSTEMS OR APPARATUS
- G02B21/00—Microscopes
- G02B21/36—Microscopes arranged for photographic purposes or projection purposes or digital imaging or video purposes including associated control and data processing arrangements
- G02B21/365—Control or image processing arrangements for digital or video microscopes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N23/00—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00
- G01N23/22—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by measuring secondary emission from the material
- G01N23/225—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by measuring secondary emission from the material using electron or ion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10056—Microscopic image
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Analytical Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Multimedia (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Optics & Photonics (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to an image processing method comprising the steps of: preparing an original image made up of two-dimensional brightness pixel values; Obtaining an entropy image having an entropy value of all pixels of the image by setting a window of a predetermined size, obtaining an entropy value of pixel values of the image in the window, and setting the entropy value as a center pixel of the window; Determining a distribution of each entropy value of the entropy image and setting an entropy threshold value of the pixel; Obtaining a black-and-white entropy-based mask image defined as 1 if the entropy value of the pixel of the entropy image is greater than or equal to the entropy threshold value and 0 if the entropy value is less than the entropy threshold value; Obtaining a filtered image of the original image with the entropy-based mask image; Determining a distribution of the filtered image and setting a brightness threshold value of the pixel; And obtaining a black-and-white image having a value of 1 when the brightness value of the pixel of the filtered image is equal to or greater than the brightness threshold value and 0 when the brightness value of the pixel of the filtered image is less than the brightness threshold value.
Description
The present invention relates to a method of dividing a monochrome image from an image in which a distinction between an object and a background such as a microscopic image is clear. This method is used for physical property evaluation based on image analysis of a material composed of a silver nanowire network, and relates to a method for deriving a black and white image segmentation for the accuracy of image analysis.
Image analysis has an important place in many engineering fields. Particularly in the field of materials processing, multifunctional films (eg transparent electrodes) or molecular sieve membranes are being produced, and the performance of the films is mainly determined by the nano- or micro-network structure and the performance can be analyzed by detecting defects . For this purpose, images photographed by optical or electron microscopes are used, and high magnification images are required because nanowires, carbon nanotubes, and zeolite crystals are used. Here, it is very important to understand the connectivity, arrangement, and orientation of network structures composed of these nano- and micro-materials. From this microscopic information, it can ultimately be used to select the desired process conditions, or it can be used to design the manufacturing process. In other words, macroscopic processing can be used to reverse the microscopic structure.
Most of the images obtained from the microscope are gray scale images. An image is a matrix composed of two-dimensional arrays, each of which is constituted by a unit of pixels, and the value represents the brightness value in the image. Image segmentation is the most basic and important step, which is the first step in image analysis, which is the process of dividing images into homogeneous parts without overlapping.
Various methods for image segmentation have been proposed until recently, for example edge extraction, partial differential approximation, fuzzy logic approach. Among them, the threshold processing is a technique which is simple and old but is still widely used in image segmentation. The bi-level thresholding is to divide the image into 0 and 1 based on the brightness value of a given pixel, for example, to distinguish between material and background in the image. This classification is classified according to whether the threshold value is based on global information or local information. In the global threshold processing, the entire image is divided by using a threshold value, and Otsu's method of selecting a valley part in the brightness histogram is representative. Local threshold processing is a method of using local information. The threshold value varies according to each area. The most representative method is the method of sauvola. Some methods can use this global thresholding and local thresholding together, which is called the two-stage method. Examples using this method are the indicator kriging method and the active contour method.
In the present invention, the present inventors have been interested in image segmentation from a microscopic image of a network structure composed of a thin object. This is mainly due to a linear feature (a higher sequence of pixels with a higher or higher intensity value than the immediate lateral neighborhood), which can be segmented by a special method called linear feature detection. Although this method has been used in various fields, the examples used in material analysis have been rare.
Global thresholding is criticized because it proceeds with ignoring local information, but it can be useful for microscopic images because the brightness of the image is greatly influenced by the state of the surface of the material. For example, in scanning electron microscopy (SEM) images, the brightness value is determined by the energy value of the secondary electron, which is determined by the structure and ratio of the surface of the material. Therefore, if we combine global threshold processing and local information appropriately, successful image segmentation can be expected.
In the present invention, entropy-based masking (EBM) is used to perform the thresholding process. Here we were able to select the pixels with more information to help rational and easy thresholding.
The present invention relates to an image processing method comprising the steps of: preparing an original image made up of two-dimensional brightness pixel values; Obtaining an entropy image having an entropy value of all pixels of the image by setting a window of a predetermined size, obtaining an entropy value of pixel values of the image in the window, and setting the entropy value as a center pixel of the window; Determining a distribution of each entropy value of the entropy image and setting an entropy threshold value of the pixel; Obtaining a black-and-white entropy-based mask image defined as 1 if the entropy value of the pixel of the entropy image is greater than or equal to the entropy threshold value and 0 if the entropy value is less than the entropy threshold value; Obtaining a filtered image of the original image with the entropy-based mask image; Determining a brightness distribution of the filtered image and setting a brightness threshold value of the pixel; And obtaining a black-and-white image of 0 if the brightness value of the pixel of the filtered image is equal to or greater than the brightness threshold value and less than 0 if the brightness value of the pixel of the filtered image is greater than or equal to the brightness threshold value.
And the original image is a microscope image.
The distribution of the entropy values is a distribution estimated from a histogram of the number of entropy values of each pixel, and the entropy threshold value is a sum of an average value and a standard deviation value of the distribution.
The step of obtaining the filtered image of the original image includes multiplying the pixel value of the original image by a value of 0 or 1 of the corresponding pixel of the monochrome entropy based mask image.
FIG. 1A illustrates entropy-based masking (EBM) of the present invention.
FIG. 1B shows the effect of the size of the entropy estimation window.
Figure 2 shows a filtered histogram of a filtered image obtained using a monochrome entropy-based mask.
FIG. 3 shows an algorithm for obtaining a divided monochrome image of the original image of the present invention, and the drawings for the right side show a series of changes according to the algorithm of the silver nanowire network image.
Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. The present invention is capable of various modifications and various forms, and specific embodiments are illustrated in the drawings and described in detail in the text. It is to be understood, however, that the invention is not intended to be limited to the particular forms disclosed, but on the contrary, is intended to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the term "comprises" or "having ", etc. is intended to specify that there is a feature, step, operation, element, part or combination thereof described in the specification, , &Quot; an ", " an ", " an "
Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the contextual meaning of the related art and are to be interpreted as either ideal or overly formal in the sense of the present application Do not.
FIG. 1A illustrates an entropy-based masking (EBM) method for obtaining a black-and-white entropy-based mask image of the present invention. FIG. 1A illustrates a histogram image as a distribution of entropy values of each pixel of the method of obtaining the entropy image of the present invention, And setting an entropy threshold value of the pixel. 1A is an SEM image of a silver nanowire network fabricated by dip coating as an example of an original image of the present invention. (B) is an entropy-estimation window, (C) shows a histogram showing an entropy distribution, and illustrates a method of setting an entropy threshold value.
The original image is an 8-bit grayscale image, and each pixel has a brightness intensity value of an integer from 0 to 255. Assuming that the values of all pixels are statistically independent, the amount of information can be expressed as the relative frequency of the brightness values in the image. The following equation (1) can be used to obtain the entropy value of the information in the area given by the calculation of entropy
Where p (i) represents the relative frequency of brightness within an entropy-estimating window, and n is the number of brightness values in the region.
As shown in FIG. 1A, a square window having a size of N around one pixel is set, where the size of N is related to the size of the object. The entropy in the window is based on the above equation (1) and is replaced by the center position of the window. If the pixel is located at the end of the image, symmetric padding is used to prevent the window from crossing the image. This is done by filling the current image with a mirror image.
Having a large entropy means that the brightness is more randomly arranged around the pixel. In other words, the change of the brightness value in the window is severe. Therefore, the value of a large entropy can be inferred that the pixel is at the boundary of matter and background. In the present invention, only pixels having a large entropy value are selected, a mask is created using the pixels, and the threshold value is used.
To select the pixels, the present invention has chosen a statistical approach. The entropy histogram is obtained by collecting all entropy values as shown in Fig. 1A (C), and then the distribution is estimated as a Gaussian distribution. Of course, you can use other distributions if necessary. As shown in FIG. 1 (C), only the pixels differing by more than the standard deviation from the average are selected for EBM, which means pixels having a high entropy of 16%.
A successful EBM depends on choosing an appropriate size entropy estimation window. The goal of EBM is to draw the boundaries between material and background. Therefore, the size of the window should be large enough to contain material and background. Due to the noise, the entropy value can become large due to the change in the brightness value even in the homogeneous part like the background. Therefore, if you size a window that is too small for a material, there is a risk that the noise portion will be caught rather than the boundary between the material and the background.
1B shows the effect of the size of the entropy estimation window. The figures in the second column show the results of the EBM in FIG. 3A according to the size of the window. The thickness of the object is about 6-9 pixels. As shown in the third column, the calculated threshold values remain almost constant after 11 by 11 magnitude. Also, the result of image segmentation becomes almost the same. However, if the size of the entropy estimation window becomes too small, the distribution of entropy becomes too much left-tailed rather than Gaussian. For example, if you are going to use a 3 by 3 window, it will only find pixels that are slightly changing, such as noise, since this is only an estimate of 9 pixels. Because EBM only uses pixels above the value of the entropy plus the standard deviation from the mean, it is rare that no pixels are selected. As a result, in terms of stability, it is recommended that the size of the entropy estimation window be larger than the size of the object.
Figure 2 shows a filtered image image using a monochrome entropy-based mask and illustrates a histogram as a distribution of the filtered image. (B) is the corresponding brightness value histogram of each image (where the black bar is the original image and the purple bar is the filtered extracted image), (C) (D) and (E) are the relative frequency of the original image and the cumulative relative frequency of the extracted image.
In the brightness histogram of the original image, the dark brightness value indicating the background dominates the overall brightness value in too much ratio. In FIG. 2 (E), in the cumulative frequency graph of the original brightness histogram, a low brightness value (brightness of about 40 or less) accounts for about 70% of the total. Therefore, it is very difficult to find appropriate threshold values from this histogram using binormal mixture analysis. In contrast, when the EBM method is used, the distribution of the relative frequency is widespread. Therefore, it can be seen that two binomial Gaussian distributions clearly represent the brightness histogram when binormal mixture analysis is performed. It was easy to select the threshold value because it was well satisfied with the condition that can be separated well.
FIG. 3 shows an algorithm for obtaining a divided monochrome image of the original image of the present invention, and the drawings for the right side show a series of changes according to the algorithm of the silver nanowire network image.
(A) is the original image, and (C) is the monochrome entropy-based mask image. (E) is a brightness distribution histogram, such as the (D) after the filtering with a mask image threshold value (in this case two threshold values (T 0, by using the T 1) the value of T 0 or less is treated as zero, and T 1 The above values were treated as 1, and the value between 0 and 1 was given by statistical interpolation.
Claims (4)
An entropy image having entropy values of all pixels of the image is obtained by setting a window having a size larger than the thickness of the wire, obtaining an entropy value of pixel values of the image in the window, and setting the entropy value as a center pixel of the window step;
Determining a brightness distribution of each entropy value of the entropy image and setting an entropy threshold value of the pixel;
Obtaining a black-and-white entropy-based mask image defined as 1 if the entropy value of the pixel of the entropy image is greater than or equal to the entropy threshold value and 0 if the entropy value is less than the entropy threshold value;
Obtaining a filtered image of the original image with the entropy-based mask image;
Determining a brightness distribution of the filtered image and setting a brightness threshold of the pixel, wherein the brightness threshold comprises a first brightness threshold and a second brightness threshold, and wherein the first brightness threshold comprises a second brightness The threshold being greater than a threshold value;
Wherein the first brightness threshold value and the second brightness threshold value are set to 1 when the brightness value of the pixel of the filtered image is greater than or equal to the first brightness threshold value and to 0 when the brightness value is less than the second brightness threshold value, 0.0 > 0 < / RTI > or 1 to obtain a monochrome image.
A method of deriving a black and white image of a wire network shape image.
Wherein the original image is a microscope image,
A method of deriving a black and white image of a wire network shape image.
Wherein the distribution of entropy values is a distribution estimated from a histogram of the number of entropy values of each pixel and the entropy threshold is a sum of the mean values of the distributions and a standard deviation value.
A method of deriving a black and white image of a wire network shape image.
Wherein the step of obtaining the filtered image of the original image comprises multiplying the pixel value of the original image by a value of 0 or 1 of the corresponding pixel of the monochrome entropy based mask image.
A method of deriving a black and white image of a wire network shape image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020150173731A KR101757071B1 (en) | 2015-12-08 | 2015-12-08 | Method of image segmentation assisted entropy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020150173731A KR101757071B1 (en) | 2015-12-08 | 2015-12-08 | Method of image segmentation assisted entropy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| KR20170067264A KR20170067264A (en) | 2017-06-16 |
| KR101757071B1 true KR101757071B1 (en) | 2017-07-12 |
Family
ID=59278436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| KR1020150173731A KR101757071B1 (en) | 2015-12-08 | 2015-12-08 | Method of image segmentation assisted entropy |
Country Status (1)
| Country | Link |
|---|---|
| KR (1) | KR101757071B1 (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109934784A (en) * | 2019-03-12 | 2019-06-25 | 中国科学院长春光学精密机械与物理研究所 | Image enchancing method, device, equipment and computer readable storage medium |
| CN111862103A (en) * | 2019-04-25 | 2020-10-30 | 中国科学院微生物研究所 | Method and device for judging cell change |
| KR102411452B1 (en) * | 2020-06-09 | 2022-06-20 | 서울대학교산학협력단 | Conductivity analysis method of conducting rod network |
| KR102663964B1 (en) * | 2022-12-09 | 2024-05-10 | 동아대학교 산학협력단 | Entropy filter implementation method and hardware device for implementing the same |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3860540B2 (en) * | 2000-12-01 | 2006-12-20 | 独立行政法人科学技術振興機構 | Entropy filter and region extraction method using the filter |
| JP2007306083A (en) | 2006-05-09 | 2007-11-22 | Konica Minolta Holdings Inc | Imaging apparatus and signal processing apparatus |
| KR101025568B1 (en) | 2009-10-12 | 2011-03-28 | 중앙대학교 산학협력단 | Apparatus and method for focusing position decision using entropy of image |
| KR101066734B1 (en) | 2010-06-21 | 2011-09-21 | 중앙대학교 산학협력단 | Method and apparatus for texture segmentation based on multi-scale entropy profile |
| KR101503606B1 (en) | 2013-10-15 | 2015-03-17 | 창원대학교 산학협력단 | Picture Quality Improvement Apparatus and Method based on Detail Information |
-
2015
- 2015-12-08 KR KR1020150173731A patent/KR101757071B1/en active IP Right Grant
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3860540B2 (en) * | 2000-12-01 | 2006-12-20 | 独立行政法人科学技術振興機構 | Entropy filter and region extraction method using the filter |
| JP2007306083A (en) | 2006-05-09 | 2007-11-22 | Konica Minolta Holdings Inc | Imaging apparatus and signal processing apparatus |
| KR101025568B1 (en) | 2009-10-12 | 2011-03-28 | 중앙대학교 산학협력단 | Apparatus and method for focusing position decision using entropy of image |
| KR101066734B1 (en) | 2010-06-21 | 2011-09-21 | 중앙대학교 산학협력단 | Method and apparatus for texture segmentation based on multi-scale entropy profile |
| KR101503606B1 (en) | 2013-10-15 | 2015-03-17 | 창원대학교 산학협력단 | Picture Quality Improvement Apparatus and Method based on Detail Information |
Also Published As
| Publication number | Publication date |
|---|---|
| KR20170067264A (en) | 2017-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101757071B1 (en) | Method of image segmentation assisted entropy | |
| CN107507173B (en) | No-reference definition evaluation method and system for full-slice image | |
| Arora et al. | Multilevel thresholding for image segmentation through a fast statistical recursive algorithm | |
| US8965123B2 (en) | System and method for processing image for identifying alphanumeric characters present in a series | |
| CN110378313B (en) | Cell cluster identification method and device and electronic equipment | |
| CN114972326A (en) | Defective product identification method for heat-shrinkable tube expanding process | |
| DE102008030874A1 (en) | Method and device for determining a contour and a center point of an object | |
| CN111462076A (en) | Method and system for detecting fuzzy area of full-slice digital pathological image | |
| CN107392946B (en) | Microscopic multi-focus image sequence processing method for three-dimensional shape reconstruction | |
| CN115330795B (en) | Cloth burr defect detection method | |
| US8000535B2 (en) | Methods and systems for refining text segmentation results | |
| CN115731400A (en) | X-ray image foreign matter detection method based on self-supervision learning | |
| CN113326846A (en) | Rapid bridge apparent disease detection method based on machine vision | |
| CN112348831A (en) | Shale SEM image segmentation method based on machine learning | |
| CN111429468A (en) | Cell nucleus segmentation method, device, equipment and storage medium | |
| CN114565607A (en) | Fabric defect image segmentation method based on neural network | |
| CN107862689A (en) | Leather surface substantially damaged automatic identifying method and computer-readable recording medium | |
| CN113379640B (en) | Multi-stage filtering image denoising method integrating edge information | |
| CN109461136B (en) | Method for detecting fiber distribution condition in mixed fiber product | |
| Khan et al. | Segmentation and quantification of activated sludge floes for wastewater treatment | |
| Salamah et al. | Enhancement of low quality thick blood smear microscopic images of malaria patients using contrast and edge corrections | |
| Matula et al. | Quantification of fluorescent spots in time series of 3D confocal microscopy images of endoplasmic reticulum exit sites based on the HMAX transform | |
| JP2010108113A (en) | Character recognition device | |
| Sarkar et al. | Image pyramid for automatic segmentation of fabric defects | |
| Kasmin et al. | Automatic Road Crack Segmentation Using Thresholding Methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| E701 | Decision to grant or registration of patent right | ||
| GRNT | Written decision to grant |