WO2012078026A1 - Method for color classification and applications of the same - Google Patents
Method for color classification and applications of the same Download PDFInfo
- Publication number
- WO2012078026A1 WO2012078026A1 PCT/MY2011/000124 MY2011000124W WO2012078026A1 WO 2012078026 A1 WO2012078026 A1 WO 2012078026A1 MY 2011000124 W MY2011000124 W MY 2011000124W WO 2012078026 A1 WO2012078026 A1 WO 2012078026A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- color
- values
- pixels
- categorized
- pixel
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- 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
- G06T2207/10024—Color image
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/6002—Corrections within particular colour systems
- H04N1/6005—Corrections within particular colour systems with luminance or chrominance signals, e.g. LC1C2, HSL or YUV
Definitions
- the present invention generally relates to visual technologies, and more particularly to a method of color classification based on RGB and HSI values of pixels and further to applications employing the method of color classification.
- color is one of the powerful descriptors. Color is one of the useful features that are widely used for identifying similar objects or performing object tracking in a surveillance application.
- US 6,757,428 discloses a color characterization method that operates to analyze each respective pixel of at least a subset of the pixels of an image object.
- the image is obtained in HSI format, or alternatively converted from another format to HSI.
- the method determines a color category or bin for the respective pixel based on values of the respective pixel.
- the color category is one of a plurality of possible color categories or bins in the HSI color space.
- the method stores information in to computer regarding the number or percentage or pixels in each of the color categories.
- the disclosed method does not use RGB values of the pixels.
- the disclosed method uses I and S values for determining black and white color while other colors are determined using H and S values.
- US 5,754,448 discloses a system and method for color characterization and transformation that obtains color data representing output of a color imaging system, and converts the color data using a color space having a white reference vector that is adjusted during the conversion.
- the white reference vector can be adjusted according to intensities of the color data being converted. Adjustment of the white reference vector serves to avoid nonuniformities for color imaging systems having different imaging bases, and thereby eliminates, or at least reduces, the amount of empirical adjustment necessary to obtain an acceptable visual match between the color imaging systems.
- the disclosed system and method uses CIELab color model.
- One aspect of the present invention provides a method of color classification of a plurality of pixels in an image.
- the method comprises the steps of acquiring the RGB values of each respective pixel from a plurality of pixels of an input image; acquiring the HSI values of each respective pixel from the plurality of pixels; providing a set of color categories, wherein each color category is defined by a combination of ranges of H and I values; determining whether the color of each respective pixel is categorized as either black or white based on the RGB values; determining whether the color of each respective pixel is categorized as either grey or dark grey based on the S and I values if it is not categorized as black or white; determining the color category of each respective pixel based on its H and I values using the provided set of color categories if it is not categorized as black, white, grey or dark grey; and storing the color categories of the plurality of pixels; thereby the plurality of pixels of the input image are represented by the stored color categories.
- Another aspect of the present invention provides an object tracking process being used in a surveillance device.
- the process comprises reading an input image; separating foreground objects from background; performing the color classification for every pixels in the foreground objects using the color classification method; obtaining the dominant color for each of the foreground objects; performing color matching on the foreground objects between current frame and previous frame; and updating the object properties (tracked) if a match is found or labeling the object as a new object if no match is found.
- the device comprises a computer executable medium for storing information and embedding algorithms; and a color classification algorithm embedded in the computer executable medium for performing the color classification.
- FIG 1 shows a typical HSI color space known in the prior art.
- FIG 2 is an exemplary graph depicting the general principle of deriving a set of color categories from H and I values in accordance with one embodiment of the present invention.
- FIG 3 is a functional flowchart illustrating the color classification in accordance with one embodiment of the present invention.
- FIG 4 shows two exemplary graphs illustrating adaptive color classification in accordance with the present invention.
- FIG 5 shows an exemplary graph illustrating that the values of one or more parameters can be varied suitable for different requirements in accordance with the present invention.
- FIG 6 is an exemplary flowchart of an object tracking process employing the color classification method of the present invention.
- HSI color space is more precise in describing colors as how humans perceive the colors. For example, when we describe the color of an object, we never say how red, green or blue it is. Instead, we describe the color, and how dark or light the object is. This is exactly what HSI color space describes colors.
- the HSI color space as illustrated in FIG 2. consists of Hue angle (0° ⁇ H
- the H component defines the purity of colors; 0° means red, 60° means yellow. 120° means green. 180° means cyan, 240° means blue, and 300° means magenta.
- the S component defines how strong a particular color will be or how much the color is polluted with white color. In a pure spectrum, colors are fully saturated.
- the I component specifies the brightness, where 0 means pure black and 1 means pure white. Following are the equations (1-3) to convert normalized R, G and B values (range from 0 to 1) to H, S and I values.
- One aspect of the present invention provides a method of color classification.
- the method classifies the colors of pixels into a number of color categories based on RGB and HSI values.
- the number of color categories can be adjusted to suit for different applications.
- a pixel or a blob (group of pixels) is classified into one color category according to its RGB and HSI values.
- the pixels with the same color category have similar color in terms of human perception.
- the hue plane can be divided into primary colour including red. green and blue, and secondary colour including cyan, magenta and yellow. If the intensity is lower than the hue plane is divided into 3 sectors which consist of dark red, dark green and dark blue. If the intensity is between Ij ow and Iu h , the hue plan is divided into 6 sectors which consist of red, green, blue, dark cyan, dark magenta and dark yellow. If the intensity is higher than Iu gt , the hue plane is again divided into 6 sectors, which consist of red, green, blue, cyan, magenta and yellow.
- the inner circle (dotted circle) is the boundary of S grey that would be categorized as grey or dark grey. It is apparent that each color category is defined by a combination of ranges of H and I values, and the full set of color categories covers all possible combinations of ranges of H and I values; thus a pixel or blob with a combination of specific H and I values can be easily classified into one of the color categories
- an image is comprised of a number of pixels or blobs, and the color classification might be needed to be performed on a portion of the pixels or blobs or all of the pixels or blobs.
- the color classification of one pixel or blob is used herein as an example to illustrate the method of the color classification of the present invention.
- the set of color categories is a default one or one specifically produced for the pixels to be classified according to the principles described above. Then, determining whether the color of the pixel or blob can be categorized as either black or white based on its RGB values. Then, determining whether the color of the pixel or blob can be categorized as either grey or dark grey based on its S and I values if it cannot be categorized as black or white. Finally, determining the color category of the pixel or blob based on its H and I values using the provided set of color categories if it cannot be categorized as black, white, grey or dark grey.
- FIG 3 there is provided a functional flowchart illustrating the color classification in accordance with one embodiment of the present invention.
- the method first determines whether the pixel is black, white or grey.
- Max(R,G,B) means the maximum values among R, G and B
- Min(R,G,B) means the minimum values among R, G and B. If Max(R,G.B) is smaller than ⁇ then it is considered as black pixel. On the other hand. If Min(R,G.B) is greater than ⁇ ,,, » then it is considered as white pixel.
- Tmin and Tnux is range from 0 to 1 , andtheir default values are 0.25 and 0.85 respectively.
- the method determines whether the pixel is grey using the Saturation value. If S is less than S ⁇ , then the pixel is further classified as grey or dark grey depending on its Intensity value. If I is less than Igrey, the pixel is classified as dark grey; otherwise, it is grey. Both Sgrey and Ig ⁇ are in the range from 0 to 1 and their default values are 0.1 and 0.5 respectively.
- each sector is having equal size ( 120° or Z60°) as illustrated in FIG 2.
- the hue plane can be further divided into smaller sectors.
- the size of each sector and the Iiow and ⁇ may vary according to the environment. By default, the values for Iio W and 1 ⁇ are 0.3334 and 0.6667 respectively.
- the number of sectors can be varied according to the intensity segment and the hue plane can be partitioned into inner ring (solid line) and outer ring.
- the value of S g r ey may change according to intensity values, which results in the change of the dotted circle size.
- the intensity axis is divided into 4 segments; in the first and second segments, the hue plane may partition into 3 sectors; in the third and forth segments, the hue plane may partition into 6 sectors.
- each sector is further divided into inner ring and outer ring.
- the size of the dotted circle (boundary for grey colour) is gradually getting smaller and smaller.
- the method of color classification of the present invention is suitable for many different applications.
- One example is color matching.
- Color matching can be done between a target object and a current object based on the categorised color instead of original color to perform object identification or object tracking.
- FIG 6 there is provided an exemplary flowchart of an object tracking process employing the adaptive color classification method of the present invention.
- the object tracking process first reads an input image, then separates foreground objects from background, perform adaptive color classification for every pixels in the foreground objects using the adaptive color classification method of the present invention, obtain the dominant color for each of the foreground objects, perform color matching on the foreground objects between current frame and previous frame, and update the object properties (tracked) if a match is found or label the object as a new object if no match is found.
- the color classification method presented here is not limited to be used in surveillance applications only.
- it can be implemented in the field of machine vision to detect faulty component based on color or in dye industry to classify colors.
- Another aspect of the present invention provides an electronic device that comprises a color classification algorithm.
- the color classification algorithm is embedded in a computer executable medium that classifies the colors of pixels into different categories, reducing the color numbers in an image.
- the adaptive color classification module reduces the requirement of computer powers for the electronic device and makes the device more tolerant to noises.
- the electronic device can be an object tracking system, a fault detection device or a color classifying device.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Spectrometry And Color Measurement (AREA)
- Color Image Communication Systems (AREA)
Abstract
The present invention provides a method of color classification of a plurality of pixels in an image. The method comprises the steps of providing a set of color categories, wherein each color category is defined by a combination of ranges of H and I values; determining the color categories of pixels based their RGB values and H and I values using the provided set of color categories. The present invention also provides an object tracking process being used in a surveillance device. The present invention further provides an electronic device that is enabled for color classification.
Description
METHOD FOR COLOR CLASSIFICATION AND APPLICATIONS OF THE
SAME
Field of the Invention
[0001] The present invention generally relates to visual technologies, and more particularly to a method of color classification based on RGB and HSI values of pixels and further to applications employing the method of color classification. Background of the Invention
[0002] In machine vision or computer vision, color is one of the powerful descriptors. Color is one of the useful features that are widely used for identifying similar objects or performing object tracking in a surveillance application.
[0003] However, in machine vision, image pixels are usually composed from Red,
Green and Blue, each measured in 8 bits. This RGB color model creates 28*3 = 16777216 potential colour categories for each pixel. When it comes to finding the matching pixels or matching blobs, it creates 167772162 matching possibility. This requires a lot of computation time and the results will be very sensitive to noise too.
[0004] US 6,757,428 discloses a color characterization method that operates to analyze each respective pixel of at least a subset of the pixels of an image object. The image is obtained in HSI format, or alternatively converted from another format to HSI. For each respective pixel, the method determines a color category or bin for the respective pixel based on values of the respective pixel. The color category is one of a plurality of possible color categories or bins in the HSI color space. As the pixels are analyzed and assigned to color categories, the method stores information in to computer regarding the number or percentage or pixels in each of the color categories. However, the disclosed method does not use RGB values of the pixels. In addition, the disclosed method uses I and S values for determining black and white color while other colors are determined using H and S values.
[0005] US 5,754,448 discloses a system and method for color characterization and transformation that obtains color data representing output of a color imaging system, and
converts the color data using a color space having a white reference vector that is adjusted during the conversion. The white reference vector can be adjusted according to intensities of the color data being converted. Adjustment of the white reference vector serves to avoid nonuniformities for color imaging systems having different imaging bases, and thereby eliminates, or at least reduces, the amount of empirical adjustment necessary to obtain an acceptable visual match between the color imaging systems. However, the disclosed system and method uses CIELab color model.
Summary of the Invention
[0006] One aspect of the present invention provides a method of color classification of a plurality of pixels in an image. In one embodiment, the method comprises the steps of acquiring the RGB values of each respective pixel from a plurality of pixels of an input image; acquiring the HSI values of each respective pixel from the plurality of pixels; providing a set of color categories, wherein each color category is defined by a combination of ranges of H and I values; determining whether the color of each respective pixel is categorized as either black or white based on the RGB values; determining whether the color of each respective pixel is categorized as either grey or dark grey based on the S and I values if it is not categorized as black or white; determining the color category of each respective pixel based on its H and I values using the provided set of color categories if it is not categorized as black, white, grey or dark grey; and storing the color categories of the plurality of pixels; thereby the plurality of pixels of the input image are represented by the stored color categories.
[0007] Another aspect of the present invention provides an object tracking process being used in a surveillance device. In one embodiment, the process comprises reading an input image; separating foreground objects from background; performing the color classification for every pixels in the foreground objects using the color classification method; obtaining the dominant color for each of the foreground objects; performing color matching on the foreground objects between current frame and previous frame; and updating the object properties (tracked) if a match is found or labeling the object as a new object if no match is found.
[0008] Another aspect of the present invention provides an electronic device. In one embodiment, the device comprises a computer executable medium for storing information and embedding algorithms; and a color classification algorithm embedded in the computer executable medium for performing the color classification.
[0009] The objectives and advantages of the invention will become apparent from the following detailed description of preferred embodiments thereof in connection with the accompanying drawings.
Brief Description of the Drawings
[0010] Preferred embodiments according to the present invention will now be described with reference to the Figures, in which like reference numerals denote like elements.
[0011] FIG 1 shows a typical HSI color space known in the prior art.
[0012] FIG 2 is an exemplary graph depicting the general principle of deriving a set of color categories from H and I values in accordance with one embodiment of the present invention.
[0013] FIG 3 is a functional flowchart illustrating the color classification in accordance with one embodiment of the present invention.
[0014] FIG 4 shows two exemplary graphs illustrating adaptive color classification in accordance with the present invention.
[0015] FIG 5 shows an exemplary graph illustrating that the values of one or more parameters can be varied suitable for different requirements in accordance with the present invention.
[0016] FIG 6 is an exemplary flowchart of an object tracking process employing the color classification method of the present invention..
Detailed Description of the Invention [0017] The present invention may be understood more readily by reference to the following detailed description of certain embodiments of the invention.
[0018] Throughout this application, where publications are referenced, the disclosures of these publications are hereby incorporated by reference, in their entireties, into this application in order to more fully describe the state of art to which this invention pertains.
[0019] Digital images or videos are often represented by the colors of pixels using the RGB color model, where each channel is measured in 8 bit. The complete representation of the color of a pixel would be 24 bits. The problem is that a family of colors like purple and the combinations close to it cannot be easily known from the 24-bit representation. For example, (255, 255, 255) means that the color of a pixel is pure white and (0, 0, 0) means that the color of a pixel is pure black. However, if you are given a pixel with (245, 39, 176), it is hard to tell the color of the pixel.
[0020] HSI color space is more precise in describing colors as how humans perceive the colors. For example, when we describe the color of an object, we never say how red, green or blue it is. Instead, we describe the color, and how dark or light the object is. This is exactly what HSI color space describes colors.
[0021] The HSI color space, as illustrated in FIG 2. consists of Hue angle (0° < H
< 360°), color Saturation (0 < S < 1) and Intensity (0 < I < 1). The H component defines the purity of colors; 0° means red, 60° means yellow. 120° means green. 180° means cyan, 240° means blue, and 300° means magenta. The S component defines how strong a particular color will be or how much the color is polluted with white color. In a pure spectrum, colors are fully saturated. The I component specifies the brightness, where 0 means pure black and 1 means pure white. Following are the equations (1-3) to convert normalized R, G and B values (range from 0 to 1) to H, S and I values.
[0022] 1 (1)
[0024]
Sf ■ COT"
Wtf» -^s + (* - ffK0 - *") (3)
[0025] where min(R.G,B) is the minimum values among R, G and B. The equation
(3) for H yields values in the interval [0°, 180°]. If B > G, then H is greater than 180° and is obtained as H = 360° - H.
[0026] One aspect of the present invention provides a method of color classification. The method classifies the colors of pixels into a number of color categories based on RGB and HSI values. The number of color categories can be adjusted to suit for different applications. A pixel or a blob (group of pixels) is classified into one color category according to its RGB and HSI values. The pixels with the same color category have similar color in terms of human perception.
[0027] Now referring to FIG 2, there is provided an exemplary graph depicting the general principle of deriving a set of olor categories from H and I values in accordance with one embodiment of the present invention. The hue plane can be divided into primary colour including red. green and blue, and secondary colour including cyan, magenta and yellow. If the intensity is lower than the hue plane is divided into 3 sectors which consist of dark red, dark green and dark blue. If the intensity is between Ijow and Iu h, the hue plan is divided into 6 sectors which consist of red, green, blue, dark cyan, dark magenta and dark yellow. If the intensity is higher than Iugt,, the hue plane is again divided into 6 sectors, which consist of red, green, blue, cyan, magenta and yellow. The inner circle (dotted circle) is the boundary of Sgrey that would be categorized as grey or dark grey. It is apparent that each color category is defined by a combination of ranges of H and I values, and the full set of color categories covers all possible combinations of ranges of H and I values; thus a pixel or blob with a combination of specific H and I values can be easily classified into one of the color categories
[0028] In reality, an image is comprised of a number of pixels or blobs, and the color classification might be needed to be performed on a portion of the pixels or blobs or all of the pixels or blobs. For the sake of simplicity, the color classification of one pixel or blob is used herein as an example to illustrate the method of the color classification of the present invention. First, acquiring the RGB values of a pixel or blob. Then, acquiring the HSI values of the pixel or blob, where the HSI values can be imported from an external source or derived from the acquired RGB values by calculation using a locally embedded algorithm. Then, providing a set of color categories, where the set of color categories is a default one or one specifically produced for the pixels to be classified according to the
principles described above. Then, determining whether the color of the pixel or blob can be categorized as either black or white based on its RGB values. Then, determining whether the color of the pixel or blob can be categorized as either grey or dark grey based on its S and I values if it cannot be categorized as black or white. Finally, determining the color category of the pixel or blob based on its H and I values using the provided set of color categories if it cannot be categorized as black, white, grey or dark grey.
[0029] Now referring to FIG 3, there is provided a functional flowchart illustrating the color classification in accordance with one embodiment of the present invention. Assuming that an input pixel is with a normalized RGB format (range from 0 to 1), the method first determines whether the pixel is black, white or grey. Max(R,G,B) means the maximum values among R, G and B, whereas Min(R,G,B) means the minimum values among R, G and B. If Max(R,G.B) is smaller than Ίααη then it is considered as black pixel. On the other hand. If Min(R,G.B) is greater than Τ,,,» then it is considered as white pixel. Tmin and Tnux is range from 0 to 1 , andtheir default values are 0.25 and 0.85 respectively.
[0030] If the pixel is neither black nor white, the method then determines whether the pixel is grey using the Saturation value. If S is less than S^, then the pixel is further classified as grey or dark grey depending on its Intensity value. If I is less than Igrey, the pixel is classified as dark grey; otherwise, it is grey. Both Sgrey and Ig^ are in the range from 0 to 1 and their default values are 0.1 and 0.5 respectively.
[0031] If the pixel still remains unclassified, then the method will take its hue and intensity values to determine its color as described above in reference to FIG 2. By default, each sector is having equal size ( 120° or Z60°) as illustrated in FIG 2. However, when an application requires a more sensitive change in colors, the hue plane can be further divided into smaller sectors. In addition, the size of each sector and the Iiow and Ι^ may vary according to the environment. By default, the values for IioW and 1^ are 0.3334 and 0.6667 respectively. For example, as shown in FIG 4(a), if the input image is bluish, then you may have smaller sectors at the blue sector, or, as shown in FIG 4(b), if the input image is dark, then you may adjust your 1^ and Ljigh to lower values such that the color classification is more sensitive to dark color.
[0032] Moreover, the number of sectors can be varied according to the intensity segment and the hue plane can be partitioned into inner ring (solid line) and outer ring. Also, the value of Sgrey may change according to intensity values, which results in the
change of the dotted circle size. As shown in FIG 5, the intensity axis is divided into 4 segments; in the first and second segments, the hue plane may partition into 3 sectors; in the third and forth segments, the hue plane may partition into 6 sectors. In the second and third segments, each sector is further divided into inner ring and outer ring. The size of the dotted circle (boundary for grey colour) is gradually getting smaller and smaller.
[0033] The method of color classification of the present invention is suitable for many different applications. One example is color matching. Color matching can be done between a target object and a current object based on the categorised color instead of original color to perform object identification or object tracking. By reducing the number of potential colors from a full range of RGB combination to a smaller number of categorized colors, it reduces the processing time and the sensitivity to image noise.
[0034] Now referring to FIG 6, there is provided an exemplary flowchart of an object tracking process employing the adaptive color classification method of the present invention. The object tracking process first reads an input image, then separates foreground objects from background, perform adaptive color classification for every pixels in the foreground objects using the adaptive color classification method of the present invention, obtain the dominant color for each of the foreground objects, perform color matching on the foreground objects between current frame and previous frame, and update the object properties (tracked) if a match is found or label the object as a new object if no match is found.
[0035] However, the color classification method presented here is not limited to be used in surveillance applications only. For example, it can be implemented in the field of machine vision to detect faulty component based on color or in dye industry to classify colors.
[0036] Another aspect of the present invention provides an electronic device that comprises a color classification algorithm. The color classification algorithm is embedded in a computer executable medium that classifies the colors of pixels into different categories, reducing the color numbers in an image. Among the potential benefits, the adaptive color classification module reduces the requirement of computer powers for the electronic device and makes the device more tolerant to noises. The electronic device can be an object tracking system, a fault detection device or a color classifying device.
[0037] While the present invention has been described with reference to particular embodiments, it will be understood that the embodiments are illustrative and that the invention scope is not so limited. Alternative embodiments of the present invention will become apparent to those having ordinary skill in the art to which the present invention pertains. Such alternate embodiments are considered to be encompassed within the scope of the present invention. Accordingly, the scope of the present invention is defined by the appended claims and is supported by the foregoing description.
Claims
What is claimed is: 1. A method of color classification of a plurality of pixels in an image, comprising the steps of:
acquiring the RGB values of each respective pixel from a plurality of pixels of an input image ;
acquiring the HSI values of each respective pixel from the plurality of pixels; providing a set of color categories, wherein each color category is defined by a combination of ranges of H and I values;
determining whether the color of each respective pixel is categorized as either black or white based on the RGB values;
determining whether the color of each respective pixel is categorized as either grey or dark grey based on the S and I values if it is not categorized as black or white;
determining the color category of each respective pixel based on its H and I values using the provided set of color categories if it is not categorized as black, white, grey or dark grey; and
storing the color categories of the plurality of pixels;
thereby the plurality of pixels of the input image are represented by the stored color categories.
2. The method of claim 1, wherein in the step of acquiring the HSI values of each respective pixel from the plurality of pixels, the HSI values are derived from the acquired RGB values by calculation using a locally embedded algorithm.
3. The method of claim 1, wherein in the step of providing a set of color categories, the set of color categories is a default one or one specifically produced for the plurality of pixels in reference to the HSI values of the plurality of pixels.
4. The method of claim 1, wherein in the step of providing a set of color categories, the ranges of H values in a hue plane are equally divided in a range of I values.
5. The method of claim 1, wherein in the step of providing a set of color categories, the ranges of H values in a hue plane are not equally divided in a range of I values.
6. The method of claim 1, wherein in the step of determining whether the color of each respective pixel is categorized as either black or white based on the RGB values, the pixel is categorized as black if Max(R,G,B) is smaller than T^, or the pixel is categorized as white if Min(R,G,B) is greater than Τ,ω,;
wherein the Max(R,G.B) means the maximum values among R, G and B; Min(R,G,B) means the minimum values among R, G and B; and Tmm and Tnux is range from 0 to 1.
7. The method of claim 1, wherein in the step of determining whether the color of each respective pixel is categorized as either grey or dark grey based on the S and I values if it is not categorized as black or white, the pixel is classified as grey or dark grey depending on its I value if S is less than S^; if I is less than Igrey, the pixel is classified as dark grey; otherwise, it is grey:
wherein both and are in the range from 0 to 1.
8. An object tracking process being used in a surveillance device, said process comprises:
reading an input image:
separating foreground objects from background:
performing the color classification for every pixels in the foreground objects using the color classification method of claim 1 ;
obtaining the dominant color for each of the foreground objects;
performing color matching on the foreground objects between current frame and previous frame; and
updating the object properties (tracked) if a match is found or labeling the object as a new object if no match is found.
9. An electronic device, comprising: a computer executable medium for storing information and embedding algorithms: and
a color classification algorithm embedded in the computer executable medium; wherein the color classification algorithm performs the color classification according to claim 1.
10. The electronic device of claim 9, wherein the electronic device is a surveillance system, a fault detection device or a color classifying device.
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
MYPI2010005905 | 2010-12-10 | ||
MYPI2010005905 MY151310A (en) | 2010-12-10 | 2010-12-10 | Method for color classification and applications of the same |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2012078026A1 true WO2012078026A1 (en) | 2012-06-14 |
Family
ID=46207364
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/MY2011/000124 WO2012078026A1 (en) | 2010-12-10 | 2011-06-22 | Method for color classification and applications of the same |
Country Status (2)
Country | Link |
---|---|
MY (1) | MY151310A (en) |
WO (1) | WO2012078026A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109003289A (en) * | 2017-12-11 | 2018-12-14 | 罗普特(厦门)科技集团有限公司 | A kind of target following fast initializing method based on color label |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020176001A1 (en) * | 2001-05-11 | 2002-11-28 | Miroslav Trajkovic | Object tracking based on color distribution |
-
2010
- 2010-12-10 MY MYPI2010005905 patent/MY151310A/en unknown
-
2011
- 2011-06-22 WO PCT/MY2011/000124 patent/WO2012078026A1/en active Application Filing
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020176001A1 (en) * | 2001-05-11 | 2002-11-28 | Miroslav Trajkovic | Object tracking based on color distribution |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109003289A (en) * | 2017-12-11 | 2018-12-14 | 罗普特(厦门)科技集团有限公司 | A kind of target following fast initializing method based on color label |
CN109003289B (en) * | 2017-12-11 | 2021-04-30 | 罗普特科技集团股份有限公司 | Target tracking rapid initialization method based on color label |
Also Published As
Publication number | Publication date |
---|---|
MY151310A (en) | 2014-05-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8055066B2 (en) | Apparatus, system, and method for skin tone detection in a CMOS image sensor | |
Gasparini et al. | Color balancing of digital photos using simple image statistics | |
EP3342151B1 (en) | Parameter adjustments based on strength change | |
EP1909484B1 (en) | Image processing device, image processing program, and image processing method | |
EP2068569B1 (en) | Method of and apparatus for detecting and adjusting colour values of skin tone pixels | |
JP2010220197A (en) | Device and method for detecting shadow in image | |
KR101030369B1 (en) | Apparatus and method for classifing image | |
EP1211638A2 (en) | Method for blond-hair-pixel removal in image skin-color detection | |
US20070211176A1 (en) | Apparatus and method for adjusting saturation image | |
Lee et al. | Color image enhancement using histogram equalization method without changing hue and saturation | |
KR101875891B1 (en) | apparatus and method for face detection using multi detection | |
US7885458B1 (en) | Illuminant estimation using gamut mapping and scene classification | |
JP5411433B2 (en) | Video classification method, system and storage medium based on video attributes | |
US20150055858A1 (en) | Systems and methods for color recognition in computer vision systems | |
JP4988872B2 (en) | Method, apparatus and program for classifying moving objects into common colors in video (Classification of moving objects into common colors in video) | |
WO2012078026A1 (en) | Method for color classification and applications of the same | |
US10430932B2 (en) | Method and an arrangement for pseudo coloring of a digital image | |
Le Moan et al. | A class-separability-based method for multi/hyperspectral image color visualization | |
Henderson et al. | An application of gaussian mixtures: Colour segmenting for the four legged league using hsi colour space | |
CN109087362B (en) | System and method for determining dominant colors in images and non-transitory computer readable medium | |
US11451713B2 (en) | Autogaining of color pattern filtered sensors to ensure color fidelity | |
Napoleon et al. | Color Image Segmentation using OTSU Method and Color Space | |
Jo et al. | Adaptive white point extraction based on dark channel prior for automatic white balance | |
US11568622B2 (en) | Color adjustments of highlighted areas | |
JP2008085995A (en) | Color correction method and image processor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11847584 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 11847584 Country of ref document: EP Kind code of ref document: A1 |