BLOOD REGIONS SEGMENTATION FOR AUTOMATIC BLOOD GROUP …€¦ · Blood Group React with anti A React with anti B React with anti D A+ Yes - Yes A- Yes - - B+ - Yes Yes B- - Yes -

International Journal of Computer Engineering and Applications, Volume XII, Issue I, Jan. 18, www.ijcea.com ISSN 2321-3469

Ajit Danti , Kumari K 16

BLOOD REGIONS SEGMENTATION FOR AUTOMATIC BLOOD

GROUP IDENTIFICATION

Ajit Danti1, Kumari K2

Dept of Computer Applications JNN College of Engineering, Shimoga

[email protected] 1 ; [email protected]

ABSTRACT:

Recently, automatic blood group identification is receiving major attention in the field of medical research. There is not enough work carried out for automation of blood group identification to help the doctors in critical situations .The blood group of an human is mainly classified into 8 groups in that 4 of them are negatives and other four are positive these 8 groups are popularly called as ABO groups .These groups are classified on the basis of reaction with the antigens .The antigens are classified into three categories namely anti-A, anti-B, anti-D.In this paper, the blood groups are detected and segmented on the basis of color & geometrical properties of the regions by using digital image processing approaches for the automation of blood group identification. In the proposed methodology the features from blood sample images are extracted like shape, area, and size of an each region in an image. Experimental result shows the efficiency of the proposed approach.

Keywords: Image-processing, Feature extraction, Segmentation, Blood group, Pre processing.

[1] INTRODUCTION

In recent days, automatic blood group identification is getting very good scope in the

medical research. Blood group classification refers to the activity which is on the basis of the

blood sample reacted with each antigen. In the context of the reaction the blood group is easily

BLOOD REGIONS SEGMENTATION FOR AUTOMATIC BLOOD GROUP IDENTIFICATION


classified. In which pre-processing steps are done for the classification of blood groups by

segmenting the regions from a given image, morphological operations are used for classifying

the blood regions based on statistical features such as area, shape, size of the segmented regions.

For detecting the blood groups classification it needs certain human efforts and it is time

consuming, the process of classifying the blood group has following steps. Initially they take

the blood sample of a patient, and is mixed with three types of chemicals namely, AntiA, Anti-

B, Anti-D. Based on that blood sample reacted with these they determine the blood groups.

Blood groups are classified into 8 groups they are, A-Positive, B-Positive, O-Positive, AB-

Positive, A-Negative, B-Negative, O-Negative, AB-Negative

For example: The sample which reacts with Anti-A and Anti-D that sample is belongs to

the A-Positive group. Like that each blood group sample has its own reaction with these three

types of chemicals .The different samples gives the different reaction with different chemicals,

and the all the same samples will gives the same result with these chemicals.

The process of manual blood group identification is time consuming in critical situation

like patients at in serious condition, to overcome this problem it needs another system to

automate this manual process. On the basis of these information’s we segment the regions from

each image by performing the pre-processing steps by using some morphological operations.

[2] LITERATURE SURVEY

In medical field lot of research work is done. For example: Lung cancer detection using

digital Image processing On CT scan Images (Anita Chaudhary and SonitSukhraj Singh,2012),

Leukemia Cancer Cell Detection using Image Processing(Ashwini Rejintal ,Aswini.N,2016),

Brain tumor detection using image processing (Amruta pramod hebli and sudha gupta,2017),

Cancer Cells Detection Using Digital Image Processing Methods (Bhagyashri G. Patil and

Sanjeev N. Jain ,2014), Image processing based abnormal blood cells detection (Deepika N.

Patil1 and Uday P. Khot,2015), Edge Preservation with Noise Reduction in Arthritis Image

(Devkant Sen and Neha Tiwari, 2015), A Study of Various Bone Fracture Detection

Techniques (Irfan Khatik, 2017), Segmentation techniques on mammograms to detect breast

abnormality (Kamalakannan.J and Rajasekhara Babu ,2016),

Analysis and Edge Detection of Lung Cancer – Survey (C. Jeya Bharathi and P. Kabilan, 2016),

Compression And Canny Edge Detection Based Methods For Image

Segmentation(S.Muthuselvi and P.Prabhu,2016), Breast Cancer Mass Detection in

Mammograms using K-means and Fuzzy C-means Clustering (Nalini Singh, Ambarish G

Mohapatra and Gurukalyan Kanungo,2011),Image Processing Based Approach for Automated

Detection of Chronic Leukemia (Neha Sharma and Dr Supreet Kaur,2016), Breast Cancer

Detection using Image Processing Techniques (PrannoyGiri and K.Saravanakumar 2017 ),

Brain Tumor Detection Using 3D Visualization (Prachi Dhikale and Madhuri Joshi,2017),

Human Skin Cancer Recognition and Classification by Unified Skin Texture and Color

Features (Shubhangi D C and Nagaraj ,2013), Cancer Cell Detection Using Mathematical

Morphology (Sivappriya T and Muthukumaran K, 2014 ), Image Segmentation and

Identification of Brain Tumor from MRI Image (Sonam S, Gavhande and S.B.Jadhav,2013),

Image Processing for Identifying Brain Tumor using Intelligent System

http://ieeexplore.ieee.org/search/searchresult.jsp?searchWithin=%22Authors%22:.QT.Anita%20Chaudhary.QT.&newsearch=true


http://ieeexplore.ieee.org/search/searchresult.jsp?searchWithin=%22Authors%22:.QT.Sonit%20Sukhraj%20Singh.QT.&newsearch=true





(Somashekhar Swamy and P.K.Kulkarni,2015), Visualization Of 3D View Of Detected Brain

Tumor And Calculation Of Its Volume (Sushma Laxman Wakchaure and Anil Khandekar,

2015), Leukemia Detection using Image Processing (V.Venmathi and et.al,2017).

[3] BLOOD SAMPLE DATABASE

The proposed method is experimented on ABO blood sample images. This dataset is

used as the benchmark database for Performance Comparisons of segmentation of regions from

a blood sample images. Basically the ABO blood groups are classified into 8 groups , in that

four of them are positive groups and remaining four are negative groups this database contains

more than 5 images of each group such as A+,B+,O+,AB+,A-,B-,O-,AB-. For the experimental

reason these images are normalized as in the size of 240 X320.These blood groups are tested

by using the three chemicals namely anti-A, anti-B, anti-D. On the basis of the samples reacted

with these chemicals the blood group will be determined as follows:

Table 1. Blood Group Reactions

Blood Group React with anti A React with anti B React with anti D

A+ Yes - Yes

A- Yes - -

B+ - Yes Yes

B- - Yes -

AB+ Yes Yes Yes

AB- Yes Yes -

O+ - - Yes

O- - - -

Sample images for both positive and negative blood groups are as shown in Figure 1 & 2.



A+ B+ O+ AB+

Fig 1: Sample images of Positive blood groups

A- B- O- AB-

Fig 2: Sample images of Negative blood groups



[4] PROPOSED METHODOLOGY

In the proposed method the blood sample images are captured from the camera under

standard lighting conditions and later normalized to a size of 160X120 to reduce the

computational cost. Blood sample contains three regions which are subjected to reaction with

three chemicals like anti-A, anti-B, anti-D. Further image is denoised to enhance the quality of

segmentation result. In this process statistical features such as area, shape and the size are

determined for segmentation of blood group regions. The proposed methodology is as shown

in the fig 3

In first step the original image is given as an input image. Basically filters are used in pre

processing for the removal of noises from an given image for the purpose of extracting the

features of an image. High pass filter is applied an image to sharpen an image. Later the

sharpened image is converted to binary image using Otsu method [8] given in the equation (1)

and (2).

t (1) iP(i)

μ1(t) = ∑

i

t (2) iP(i)

μ2(t) = ∑

i

Ac {w|w ∉ A }

Where μ1 μ2 are mean values, t is threshold value, i is an image p and q are high probability values.

(3)

Fig 3.Block diagram

Input image

Pre - processing

Feature

extraction

Segmentation



w=pixel value of input image

A= set.

Opening process [8] is performed to remove unwanted regions from an image and also it removes all

connected components that have fewer than P pixels from the binary image using equation (3).

[5] EXPERIMENTAL RESULTS

In this work, experimentation is carried out on blood sample images containing eight blood

groups as given in Table 1. Basically the blood groups are classified into 8 categories such as

A+,B+,O+,AB+,A-,B-,AB-,O-. The data set contains 10 images of each blood group sample. The

segmentation by the proposed method is found to be more efficient in terms of speed and accuracy.

Each of the segmented region is then searched for the features using areas, size and shape to recognize

the blood groups in it.

The proposed approach is implemented using MATLAB software R2016b(9.1.0.441655). on

Core i3 @ 2.40GHz. The test images are expected to contain the different size of the regions which

are on the basis of the blood groups. The different blood group images have been segmented

successfully. In the experimental results, 158 images are properly segmented out of 168 images leading

to success rate is 94.04% (approximately). The average time taken to segment regions is about 0.178

second. The experimental results are shown in the below figures.

Original image Segmented image Final output



Fig4: Experimental results on A+ve images




Fig5: Experimental results on A-ve images




Fig6: Experimental results on B+ve images



Original Image Segmented image Final output

Fig7: Experimental results on AB+ve images





Fig8: Experimental results on O+ve images



Fig9: Experimental results O-ve images




Fig10: Experimental results on AB-ve images




Fig11: Experimental results on B-ve images



[6] CONCLUSION

In this paper, an effective segmentation of the blood regions is done using morphological

operations. The proposed approach exhibited 94.04% accuracy for our dataset. In future

by using these segmented images automation of the blood groups classification will be

carried in the future work.

REFERENCES

[1] Ashwini Rejintal and Aswini.N, Leukemia Cancer Cell Detection using Image Processing,

International Journal of Advanced Research in Electrical, Electronics and Instrumentation

Engineering, vol-06, Page No:34-38, Issue 06,ISSN: 2278 – 8875 , July 2016.

[2] Anita Chaudhary and Sonit Sukhraj Singh, Lung cancer detection International Conference14-15,

Page No:142-146, Issue 14 , ISBN: 978-0-7695-4817-3, September 2012.

[3] Amruta pramod hebli and sudha gupta, Brain tumor detection using image processing: A survey,

International Journal of Industrial Electronics and Electrical Engineering,vol-5, Page No:41-44,

Issue 1, ISSN: 2347-6982, Jan 2017.

[4] Bhagyashri G.Patil and Sanjeev N. Jain ,Research on Cancer Cells Detection, International Journal

of Latest Trends in Engineering and Technology (IJLTET), vol-3, Page No:45-49, Issue 04, ISSN:

2278-621X , march 2014.

[5] Deepika N. Patil and Uday P. Khot, Image processing based abnormal blood cells detection,

International Journal of Technical Research and Applications,Page No:37-43, Issue 31, e-ISSN:

2320-8163, September 2015.

[6] Devkant Sen and Neha Tiwari, Edge Preservation with Noise Reduction in Arthritis Image,

International Journal for Scientific Research & Development, vol-2,Page No:371375,Issue 12, ISSN

(online): 2321-0613,2015.

[7] Rafael C. Gonzalez, Richard E.Woods, Digital image processing, 2edition, Pearson,

[8] Irfan Khatik, A Study of Various Bone Fracture Detection Techniques, International Journal Of

Engineering And Computer Science,vol-6, Page No.21418-21423, Issue 05 , ISSN:2319-7242, may

2017.

[9] C. Jeya Bharathi and P. Kabilan, Analysis and Edge Detection of Lung Cancer – Survey,

International Journal on Recent and Innovation Trends in Computing and Communication,vol-

04,Page No:390-392,Issue 05, ISSN: 2321-8169,may 2016.

[10] Kamala kannan.J and Rajasekhara Babu, Segmentation techniques on mammograms to detect breast

abnormality, International Journal of Pharmacy & Technology,vol-8, Page No:16089-16099 , Issue

25, ISSN: 0975-766X , Aug 2016.

[11] S.Muthuselvi and P.Prabhu, Compression And Canny Edge Detection Based Methods For Image

Segmentation, International Journal of Advanced Research Trends in Engineering and Technology,

Page No:63-67, Issue 20, ISSN 2394-3785, April 2016.

[12] Nalini Singh, Ambarish G Mohapatra and Gurukalyan Kanungo, Breast Cancer Mass Detection in

Mammograms using K-means and Fuzzy C-means Clustering, International Journal of Computer

Applications (0975-8887), vol-22–No2 ,Page No:15-21,Issue 15 , May 2011

[13] Neha Sharma and Dr Supreet Kaur , Image Processing Based Approach for Automated Detection

of Chronic Leukemia, International Journal of Innovative Research in Computer and





Communication Engineering, vol-04, Page No:15038-15045, Issue 08, ISSN(O): 2320-9801, ISSN

(P): 2320-9798, Aug 2016.

[14] Parul Parmar and Vinay Thakur, A Review on Tumor Detection in Medical Images, International

Research Journal of Engineering and Technology (IRJET),vol-04 ,Page No:2714-2717 ,Issue 05, e-

ISSN: 2395 -0056,p-ISSN: 2395-0072, may 2017.

[15] Prachi Dhikale and Madhuri Joshi, Brain Tumor Detection Using 3D Visualization, International

Journal of Computer Science and Network, vol-06,Page No:170-177, Issue 2, ISSN (Online) : 2277-

5420, April 2017.

[16] Sergios Theodoridis Koutroumbas, Pattern Recognition 4th Edition, Page No:1-711.

[17] Shubhangi D C and Nagaraj ,Human Skin Cancer Recognition and Classification by Unified Skin

Texture and Color Features, IOSR Journal of Computer Engineering,vol-12, Page No.42-49, Issue

4,e-ISSN: 2278-0661, p-ISSN: 2278-8727, Jul-Aug 2013.

[18] Sivappriya T and Muthukumaran K, Cancer Cell Detection Using Mathematical Morphology,

International Journal of Innovative Research in Computer and

Communication Engineering, vol-2, Page No:3717-3725,Issue 1, ISSN(O): 2320-9801, March 2014

[19] Sonam S, Gavhande and S.B.Jadhav, Image Segmentation and Identification of Brain Tumor from

MRI Image, International Research Journal of Engineering and Technology,vol-02, Page No:167-

170,Issue 02,e-ISSN: 2395-0056,p-ISSN: 2395-072, May 2013.

[20] Somashekhar Swamy and P.K.Kulkarni, Image Processing for Identifying Brain Tumor using

Intelligent System , International Journal of Innovative Research in Science, Engineering and

Technology, vol-04, Page No:10937-10943, Issue 11, ISSN(O): 23198753, November 2015.

[21] Sushma Laxman Wakchaure and Anil Khandekar, Visualization Of 3D View Of Detected Brain

Tumor And Calculation Of Its Volume, International Journal of Technical Research and

Applications, vol-3, Page No:120-126,Issue 6, November-December, 2015.

[22] V.Venmathi and et.al, Leukemia Detection using Image Processing , International Journal for

Scientific Research & Development, vol-5,Page No:804-808, Issue 01, ISSN (O): 2321-0613, 2017.

[23] https://medlineplus.gov › Medical Encyclopedia.

[24] ]https://homehealth-uk.com › ... › Blood Group Type Test ABO and Rhesus (D)

https://medlineplus.gov/

https://medlineplus.gov/

https://homehealth-uk.com/





BLOOD REGIONS SEGMENTATION FOR AUTOMATIC BLOOD GROUP …€¦ · Blood Group React with anti A React with anti B React with anti D A+ Yes - Yes A- Yes - - B+ - Yes Yes B- - Yes -

Documents