Comparative Analysis of Malaria Red Blood Cell Image Segmentation

Comparative Analysis of Malaria Red Blood Cell Image Segmentation

T. Narmatha

M.Phil Research Scholar, Department of Computer Science, Periyar University,

Salem-636011.

Dr. I. Laurence Aroquiaraj

Assistant Professor, Department of Computer Science,

Periyar University, Salem-636011.

Abstract — Malaria in humans can be caused by a number of different parasites the most dangerous, and the one which is responsible for over 90% of the worldwide deaths from malaria, is Plasmodium falciparum. This Paper presents a method to Infected Malaria Parasites in Red Blood Cell (RBC). The Blood Test used to evaluate the overall health and diagnose a wide range of disorders. Our goal is to identify the Malaria Infected Red Blood Cells to use the Microscope Images for research. This study has proposed an unsupervised gray Image Segmentation of Malaria disease using Moving K-Means Cluster algorithm, FCM algorithm and Watershed Algorithm. It has been using on blood sample Microscopy Images of both types of Malaria Red Blood Cells and Normal Blood Sample Images with the aim of obtaining fully segmented abnormal and normal Red Blood Cell. The Improvement need to be done for both segmentation and overlapped cell handling to obtained better result in the future.

Keywords–Red Blood Cells, K-Means, Watershed, Fuzzy c means and Image Segmentation

I.INTRODUCTION

Most of the diseases are caused due to the blood. Malaria is one of the dangerous diseases. The detection of the malaria Blood cells are generally analyzed manually with the help of Microscope. The analysis of blood cells Images by using the different Image Processing Techniques. This paper also presents the approach for detection of the Infected Red Blood Cell in the available blood cell Image for the Evaluation of Malaria disease[8]. The Image segmentation technique is very useful for the analysis of the Infected Blood Cells Image.We present an approach to automatically detect malaria parasites in unstained blood droplets. Majority of automated image analysis algorithms are designed to detect parasites in stained cells [7]. Related image processing algorithms for the automated detection of malaria cells are applied on stained cells [5].

1. STEPS FOR PROPOSED METHODOLOGY

   In this paper, the proposed method is focused on image Processing Procedure to obtain the segmented Red Blood Cell of affected Images.

   Image Acquisition

   Image Pre- Processing

   Image Segmentation

   K Means

   Fuzzy C Means

   Watershed

   Fig1. Segmentation Methods III.FLOW DIAGRAM

   Image Acquisition

   Image Enhancement

   Image Representation

   Image Restoration

   Image Segmentation

   Image compression

   Fig2. Image Preprocessing

   1. Image acquisition:

      For detecting Malaria, microscopic images are acquired.

      These images are taken from net. This stage includes image pre processing.

      Fig3. Original Image

   2. Image pre processing:

      This is a pre-process of an Image Sequence before feeding into the segmentation process. There are various types of noise in Image processing. We have used salt and pepper noise. And after to using Median filter [3]. In digital Image processing, removing the noise is one of the preprocessing techniques.

   3. Image segmentation:

      The next stage deals with Image segmentation. Partitions an Input Image into foreground and background region. There are various approaches for segmentation. We have used three types of algorithms.

      1. K-Means 2.Fuzzy C Means 3. Watershed

   1. K- Mean Algorithm

      K clusters are formed by partitioning the dataset and the data points are randomly assigned to the Clusters resulting in clusters that have roughly the Same number of data points.

      • For all data point the distance from the data point to each cluster is calculated.

      • Leave the data point where it is only if it closes to its own cluster. If the data point is not close to its own

        cluster, shift it into the closest cluster.

        • Repeat the above step until a complete pass through all the data pointsresults in no data point moving from one cluster to another cluster. On this point the clusters are stable and the clustering process ends.

   2. Fuzzy C Means Algorithm

      1. First the initial fuzzy partition matrix is generated and the initial fuzzy cluster centers are calculated.

      2. In each step of iteration the cluster center and the membership grade point are updated and the objective function is minimized to find the best location for the cluster.

      3. Improved FCM is proposed cluster technique. It is used to solve the minimal distance.

      4. The process stops when the maximum number of iteration is reached or when the objective function improvement between two consecutive iteration is less than the minimum amount of data specified.[8]

   3. Watershed Algorithm

      The Watershed Transform is a unique technique for

      Segmenting digital images that uses a type of region

      Growing method based on an image gradient. The concept of Watershed Transform is based on visualizing an image in three dimensions: two spatial coordinates versus gray levels. In such a topographic interpretation, we consider three types of points:

      • Points belonging to a regional minimum.

      • Points at which a drop of water , if placed at the location of any of those points, would fall with certainty to a single minimum.

      • Points at which water would be equally likely to fall to more than one such minimum.

   For a particular regional minimum, the set of points Satisfying condition (B) is called the catchment basin or watershed of that minimum. The points satisfying condition

   (C) form crest lines on the topographic surface and are termed divide lines or watershed lines. The principal objective of segmentation algorithms based on these concepts is to find the watershed lines.

   1. EVALUATION PARAMETERS

      Having segmentation evaluation measures is an efficient way to analyze the performance of existing and future algorithms. Segmentation evaluation metrics can be divided into boundary based and region based methods. Before one gets to know the performance of an algorithm, knowing comprehensively the definitions of these metrics is Inevitable. Various performance parameters used for Evaluations of image segmentation are as follows [10].

      Table 1. Evaluation Measurements Formula

      S.No	Type	Description
      1.	RMSE	= 1 ( )2 =1
      2.	SNR	() = 20 ()) ( ()
      3.	PSNR	2 = 1010
      4	MAE	= 1 |(, ) (, )| =1 =1

      1. RESULT AND DISCUSSION

         The Proposed technique is applied to 22 Images of blood Cell. Images are taken from the Google. The Images are in JPG format then Image is used for contrast enhancement then the Image is segmented by K-Means, Fuzzy c Means and Watershed are compared. The result shows that the adaptive watershed algorithm gives better segmentation result as compared to k means and FCM. To obtain the RMSE, SNR, PSNR and MAE values n the paper.

         Table 2. Accuracy of Measurements

         Algorithm	RMSE	SNR	PSNR	MAE
         k-mean	143.623	-18.5465	4.91367	104.5448
         FCM	107.52	-11.966	7.581749	82.8674
         Watershed	19.90772	-0.99487	22.62104	0.983849

         160

         140

         120

         100

         80

         60

         K MEANS

         FCM WATERSHED

         40

         20

         0

         -20

         RMSE SNR PSNR MAE

         Fig.4 Accuracy of Measurements

         MAE

         120

         100

         80

         60

         MAE

         Original Images	K-Means	Fuzzy c means	Watershed
         Image 1
         Image 2
         Image 3
         Image 4
         Image 5

         Fig 6. Sample Output for Red Blood Cells

      2. CONCLUSION

In our Research we have identified affected RBC from blood cell. Image processing techniques are implemented in order to get more accurate results. Using this (RMSE,PSNR,SNR and MAE) four Evaluation Measurement, we found that Watershed Algorithm is the best segment to identify the Malaria affected Red Blood Cell.

K MEANS

FCM WATERSHED

40

20

0

Fig 5. Accuracy of Mean square Error

REFRENCE:

1. J.M.Sharif, M.F.Miswan, M.A.Ngadi, Md Sah Hj, Red Blood Cell Segmentation Using Masking and Watershed Algorithm: A Preliminary Study,International Conference onBiomedicalEngineering(ICoBE) February 2012.

2. Pooja Lepcha, Worawut srisukkham,Li Zhang, alamgir Hossain,Red Blood Based Disease Screening using Marker Controlled Watershed Segmentation and post- Processing,IEEE International Conference 2014.

3. LalitB.Damahe,R.K.Krishna,N.J.Janwe,Dr.Thankur Nileshsingh v.,Segmentation Based Approach to Detect Parasites and RBCs in Blood Cell Images,Published by Research Publications,Chikhli,India.

4. Nural Ahafikha Noor Rashid,Mohd Yusoff Mashor, Rosline Hassan,Unsupervised Color Image Segmentation of Red Blood Cell for Thalassemia Disease,IEEE International Conference 2015.

5. Ms.Snehal,Suryawanshi,Prof.V.V.Dixit, Improved Technique for Detection of Malaria Parasites within the Blood Cell Images, International Journal of Scientific & Engineering Research, Volume 4,Issue7,July 2013

6. Pawan Agrawal,Pradipti Verma,Automated Detection and Counting of Red Blood Cell usingImageProcessingTechniques,.International Journal of Scientific research and management (IJSRM) volume 3 issue 4 April2015.

7. Dr.T.Karthikeyan,N.Poornima,Microscopic Image Segmentation using Fuzzy C Means for Leukemia Diagnosis,IJARSET vol.4,Issue 1,January 2017.

8. Niket Amoda, Ramesh K Kulkarni,Image Segmentation and Detection using Watershed Transform and Region Based Image Retrieval,IJETTCS, Volume 2,Issue 2, March-April 2013.

9. Neha vyas, Mrs.Megha Singh,Improved Fuzzy C Means Clustering for Complete Blood Cell Segmentation,IJDACR,Volume 4,Issue 7,February 2016.

10. NirmalPatel,RajivKumar,Image Segmentation and Performance Evaluation,IJRASET,Vol 2, issue 1, September 2014.

11. Ravikumar, Munish Rattan ,Analysis of Various Quality Metrics for Medical Image Processing,IJARCSSE, Volume 2, Issue 1,November 2012.

12. Hany A.Elsalamony,Healthy and unhealthy red blood cell detection in human blood smears using Neural Networks,2016 Published by Elsevier Ltd.

    .

13. SumeetChourasiya, GUsha Rani,Automatic Red Blood Cell Counting Using Watershed Segmentation,IJCSIT,Vol.5(4),2014.

14. Joost Vromen,Brendan McCane,"Red Blood Cell Segmentation from SEM Images,IEEE International Conference 2013.

15. CeciliaDiRuberto, Andrew Dempster,Shahid Khan,Bill Jarra,Automatic Thresholding of Infected Blood Images Using Granulometry and Regional Extrema, IEEE International Conference 2000.