Volume 5, Issue 3, March 2015
ISSN: 2277 128X
International Journal of Advanced Research in
Computer Science and Software Engineering
Research Paper
Available online at: www.ijarcsse.com
Special Issue: E-Technologies in Anthropology
Conference Held at Bon Secours College for Women, India
Removal of Color Blindness Using Image Segmentation
R. Ranjani,
Research Scholar,
Department of Computer Science,
Mother Teresa Women‟s University, Kodaikanal
Dr. M.P. Indra Gandhi,
Assistant Professor (SG),
Department of Computer Science,
Mother Teresa Women‟s University, Kodaikanal
Abstract- Color blindness deficiency is the inability or decreased ability to see the color, under normal lighting
conditions. There is no actual blindness but there is a deficiency of color vision. Color blindness is of many types like
Red-Green, Blue-Yellow etc.., In this paper we proposed a new method that the users used the histogram thresholding
to create a green colour mask which when they applied on the original image.
KEY WORDS: Color Blindness, LMS Value, Color band, Threshold, Green Mask, Modify Image.
I. INTRODUCTION
Color blindness is a color vision problem where person is deficient to recognize colors like red, green and blue. There are
different causes of colour blindness. For the vast majority of people with deficient colour vision the condition is genetic
although some people become colour blind as a result of other diseases such as diabetes and multiple sclerosis or they
acquire the condition over time due to the aging process, medication etc. To see anything at all we need some tiny little
helpers inside our eyeballs, the so called photoreceptors. There are two different types of them: rods and cones. Both of
them are sitting on the retina and pass information of light on to our brain. There are about 120 million rods which are
very sensitive to light but not to color. The cones are the photoreceptors which are responsible for our color vision. There
are three types of cones are:
 S-cones: sensitive to short wavelength light with a peak at ca. 420nm (blue)
 M-cones: sensitive to medium wavelength light, peak at ca. 530nm (green)
 L-cones: sensitive to long wavelength light, peak at ca. 560nm (red)
This type of color blindness is usually a sex-linked condition. The genes that produce photo pigments are carried on the
X chromosome; if some of these genes are missing or damaged, color blindness will be expressed in males with a higher
probability than in females because males only have one X chromosome (in females, a functional gene on only one of the
two X chromosomes is sufficient to yield the needed photo pigments).
There are different types of color blindness
 Monochromacy: There is no cone or only one type of cone present at retina of eye. It is called Monochromacy.
 Dichromacy: There are only two types of cones present at retina of eye. It is called Dichromacy. Dichromacy is
again of three types according to missing cone: Protanopia: L-cone missing is called protanopia. So the person suffered from protanopia is unable
to see red color. This is called „Red blindness‟.
 Deuteranopia: M-cone missing is called Deuteranopia. So the person suffered from deuteranopia
is unable to see green color. This is called „Green blindness‟.
 Tritanopia: S-cone missing is called Tritanopia. So the person suffered from tritanopia is unable
to see blue color. This is called „Blue blindness‟.
 Anomalous Trichromacy: In anomalous trichromacy all three cones are present but one of these cones
perceives color slightly. It is of three types depending upon which cone is working improperly: Protanomaly: Defective of L cones and sensitivity to red hue is lower. This is called „Red
Weakness‟.
 Deuteranomaly: Defective of M cones and sensitivity to distinguish red and green hue is lower.
This is called „Green Weakness‟.
 Tritanomaly: Defective of S cones and sensitivity to distinguish blue and yellow hue is lower.
This is called „Blue Weakness‟.
 Red-green color blindness Red-green color blindness is a type of deficiency of color vision where sensing of
red and green color is weak. Red-green color blindness term is used for Protanopia, Deuteranopia,
Protanomaly and Deuteranomaly.
 Blue-yellow color blindness Blue-yellow color blindness is a type of deficiency of color vision where sensing
of blue and yellow color is weak. Blue-yellow color blindness term is used for Tritanopia, Tritanomaly.
© 2015, IJARCSSE All Rights Reserved
Page | 201
Ranjani et al., International Journal of Advanced Research in Computer Science and Software Engineering 5 (3),
March- 2015, pp. 201-204
II. PREVIOUS WORK
RGB to HSV conversion: As process starts first web contents are extracted from the websites and then out of these
contents some images that are to be transformed are selected and saved. After saving these extracted images are passed
through the colour transformation process, by which unrecognized colours are transformed to recognizable colors to the
color blind person. This research focuses on the red-green colour vision deficient. Transformation process result as red is
transformed to yellow and green is transformed to blue and blue is remaining same [2]. Green color‟s range is 120°
because its hue value comes between 60° to 180°. Blue color‟s range is 180° to 300° because its hue value comes
between 180° to 300° [4]. Green Ratio = (Hue - 60) / Green Range. RelativeBlue = GreenRatio x BlueRange Hue value
after the transformation process is Hue = 180 + BlueRatio After transformation hue value is divided by 360° for HSV to
RGB conversion. Hue = Hue / 36.Gradient map method is an approach that is able to indicate regions that encounter the
accessibility problem for colorblind viewers, the regions contain information that may not be well perceived by color
blind This method can be applied in different scenarios, such as checking the accessibility of designed images and to
help designers to avoid the accessibility problem by making changes on the image[5]. Dalton Method performs that,
different color image has been found useful to visually inspect perceived colour difference and additionally to build
colour remapping methods. Two colour transformation methods are presented [1]. This research focused on two types of
dichromacy (protanopy and deuteranopy).In this some methods are described as Image simulation, color transformation
using color difference, color transformation using color difference scaling, color transformation using red/green scaling.
LMS space plane is defined as: αL + βM + γS = 0.The basic idea is first to find the LMS values of the RGB (red-greenblue) image using some conversion matrix. Then a conversion is made to delete the information associated with the loss
of any of the cone type to get the modified LMS values L‟M‟S‟.Then reverse transformation is done on the L‟M‟S‟
values to get the R‟G‟B‟ values. Now R‟G‟B‟ values represent that how the specific color RGB is perceived by a color
blind person. When this operation is done for all the pixels, the image is converted. This linear transformation can be
achieved by a matrix multiplication [8].For each pixel do Gamma correction [R, G, B]= [R/255, G/255, B/255]^2.2
Principal Component Analysis (PCA) technique that takes a collection of data and transfer in that manner the new data
has a given statistical properties. Seperating confusion line method in color space is using color segmentation for
protanopia and deuteranopia. In the huge part of the image is used CVD making a confusion line map, 512 virtual boxes
in RGB color space.After that classify the confusion line check the region place that is belong to confusion line or not
that is calculate by the seed points and the histogram. It performs the color transformation in CIE Lab color spaces [6].
Color Blind Plate(CBP) The image segmentation and pattern recognition to color blindness plate. The CBP is wellknown satisfactory way of testing the degree of color blindness happened in the human visual system. The image of CBP
is very complex. It includes not only the colors but also the disconnected size-varied dots. It is very difficult by using a
conventional machine vision algorithm to recognize the meaningful pattern (ex: figure) from such a CBP image.1)
Passive Process : CBP Segmentation 2) Active Process: CBP Recognition[3].CCPS is a measure of the percentage of
pixels in an image that change color category when viewed by individuals with dichromatic color vision. RCCPS is a
measure of how many pixels change color category in each color category. CCPS provides an overview of how much
pixel change occurs, RCCPS provides details about which Color categories undergo changes along with how much per
category. Both the CCPS and RCCPS measures rely on image transforms as part of the process of generating the
measures [7].
𝑆ℎ𝑖𝑓𝑡𝑒𝑑 𝑃𝑖𝑥𝑒𝑙𝑠
𝐶𝐶𝑃𝑆 =
∗ 100
𝑇𝑜𝑡𝑎𝑙 𝑃𝑖𝑥𝑒𝑙𝑠
(𝐶𝑃𝑂 − 𝐶𝑃𝑇)
𝑅𝐶𝐶𝑃𝑆 =
∗ 100
𝑇𝑜𝑡𝑎𝑙 𝑃𝑖𝑥𝑒𝑙𝑠
III. METHODOLOGY
In this paper we propose a new technique to remove color blindness to make image visible to the color deficient person.
This approach has following processes:Image simulation
Find the LMS values of the RGB (red-green-blue) image using some conversion matrix. Then a conversion is made to
delete the information associated with the loss of any of the cone type to get the modified LMS values L‟M‟S‟. Then,
reverse transformation is done on the L‟M‟S‟ values to get the R‟G‟B‟ values. Now R‟G‟B‟ values represent that how the
specific color RGB is perceived by a color blind person. When this operation is done for all the pixels, the image is
converted. This linear transformation can be achieved by a matrix multiplication.
Thresholding
Thresholding is the process of extracting all the pixels in an image that lie within a specific range of colors. Color
information of each pixel in an image is typically represented as a point in a 3 dimensional space. In color images each
pixel is characterized by the three RGB values. Then to construct a 3D histogram and the basic procedure is analogous to
the method used for one variable. Histograms plotted for each of the color values and threshold points found.
Creating green mask
Using the threshold value the green region of an image can be masked.
Modifying image
After the masking, the particular region of an image is converted to purple using matrix value. Modifying matrix (M) is
given as
© 2015, IJARCSSE All Rights Reserved
Page | 202
Ranjani et al., International Journal of Advanced Research in Computer Science and Software Engineering 5 (3),
March- 2015, pp. 201-204
0.5 0.1 0
0
0
0
0.9 0 0.1
This matrix is multiply with each pixel on the original image.
IV. EXPERIMENTAL RESULTS
Figure 1:Original Image
Figure 2: Image Seen by person
suffering from Deutenophia
Figure 3: Modified image
Figure 1: Original Image, Figure 2: Image Seen by person suffering from Deutenophia, Figure 3: Modified Image shows
that the process is successful in modifying the images for colour blindness. The colour confusion of green is clearly
solved. This process takes 1.972 seconds to execute.
V. CONCLUSION
The proposed work shows the accurate results of the images for the color blindness. This procedure is very useful for
color blindness. Applying the morphological process based on the histogram value the original image is converted into
modified image. The green color has been converted into purple color as shown in Figure 3.
VI. FUTURE SCOPE
This work can be extended for various shades of green color and protanopia, tritanopia with blue and yellow color.
REFERENCES
[1]
J.B. Huang, Y.C. Tseng, S.I. Wu, and S.J. Wang, “Information preserving color transformation for protanopia
and deuteranopia,” IEEE Signal Processing. Lett., vol. 14, no. 10, Oct,2007.
[2]
Kim, Y. K., Kim, K. W. and Yang, X., “Real Time Traffic Light Recognition System for Color Vision
Deficiencies”, Mechatronics and Automation (lCMA), International Conference, pp. 76-81, 2007
[3]
Jacek Ruminski, Jerzy Wtorek, Joanna Ruminska, Mariusz Kaczmarek, Adam Bujnowski, Tomasz Kocejko,
and Artur Polinski, “Color Transformation Methods for Dichromats”, Human System Interactions (HSI), 2010
3rd Conference, pp: 634 - 641, 2010
[4]
Siew-Li Ching and Maziani Sabudin, “Website Image Colour Transformation for the Colour Blind”, Computer
Technology and Development (ICCTD), 2010 2nd International Conference, pp. 255 - 259, 2010
[5]
J.P.Srividhya, P.Sivakumar and Dr.M.Rajaram , “ The color blindness removal technique in image by using
gradient map method”, Signal Processing, Communication, Computing and Networking Technologies
(ICSCCN), 2011 International Conference, pp: 24 - 29, 20
© 2015, IJARCSSE All Rights Reserved
Page | 203
Ranjani et al., International Journal of Advanced Research in Computer Science and Software Engineering 5 (3),
March- 2015, pp. 201-204
[6]
Jinsan Park, Jongho Choi, Dongil Han “Applying enhanced confusion line Color transform using Color
segmentation for mobile applications”, IEEE ,2011
[7]
Mike Bennett and Aaron Quigley Imaging, “A Method for the Automatic Analysis of Color Category Pixel
Shifts During Dichromatic Vision”, Visualization & Graphics Lab Systems Research Group, 2012.
[8]
Ruchi Kulshrestha, R.K. Bairwa, “Removal of Color Blindness using Threshold and Masking”, International
Journal of Advanced Research in Computer Science and Software Engineering, Volume 3, Issue 6, June 2013
© 2015, IJARCSSE All Rights Reserved
Page | 204