Sie sind auf Seite 1von 4

2011 IEEE 7th International Colloquium on Signal Processing and its Applications

Image Enhancement and Segmentation Using Dark


Stretching Technique for Plasmodium Falciparum
for Thick Blood Smear
N.S.M.M. Hanif, M. Y. Mashor, Z. Mohamed
School of Mechatronic Engineering, Department of Medical Microbiology and Parasitology,
Universiti Malaysia Perlis, School of Medical Science,
Kampus Ulu Pauh, Health Campus, University Sains Malaysia,
02600 Ulu Pauh, Perlis 16150 Kubang Kerian, Kelantan, Malaysia.
Email: mira_safiah@yahoo.com Email: zeehaida@kck.usm.my

Abstract- This paper present the results of applying dark problems can hide and cause difficulty to interpret the
stretching technique to enhance and segment the Plasmodium important information in malaria slide image. Contrast is one
Falciparum based on thick blood smear images. Image of the factors that influenced the accuracy of interpretation of
enhancement is the process to improve the quality (clarity) of disease based on medical images.
images for human viewing. Removing blurring and noise,
Contrast enhancement technique plays an important role to
increasing contrast, and revealing details are examples of
enhancement operations. Reducing the noise and blurring and enhance the quality and contrast of medical images [4].
increasing the contrast range could enhance the image. The Several previous studies proved that contrast enhancement
original image might have areas of very high and very low techniques are capable to enhance the contrast of medical
intensity, which mask details. Segmentation is to cluster pixels images [4, 5]. Therefore, this paper proposed to use contrast
into salient image regions, i.e., regions corresponding to enhancement technique based on dark stretching technique to
individual surfaces, objects, or natural parts of objects. enhance and segment the Malaria Slide Images in Plasmodium
Falciparum for thick blood smear.
Keywords— Plasmodium Falciparum, Image Enhancement, Dark
Stretching II. METHODOLOGY
Currently there are two tests to detect the malaria parasite;
I. INTRODUCTION thick (Figure 1) and thin smears (Figure 2). The thick blood
Malaria is a common and life threatening disease in many smear are more useful for detecting and counting the parasites,
tropical and subtropical areas [1]. It is currently endemic in while thin blood smear helps doctors to discover what species
over 100 countries, which are visited by more than 125 of malaria is causing the infection. This paper will focus on
million international travellers every year [1]. Each year many thick blood smear.
international travellers fall ill with malaria while visiting
countries where the disease is endemic and well over 10 000
are reported to fall ill after returning home. However, under
reporting means that the real figure may be as high as 30 000
[1]. Research has shown that malaria is caused by four species
of the genus malaria parasites Plasmodium (i.e. P.
Falciparum, P. Malariae, P. Vivac and P. Ovale) [2].
Plasmodium is actually a small, single cell blood organism or
protozoan which originated from a species of mosquito called
Anopheles [3]. Anopheles mosquito bites can spread
Plasmodium into the human blood stream. Plasmodium Figure 1.Plasmodium Falciparum in Thick Blood Smear
Falciparum produces the most several symptoms and
responsible for most malaria deaths.
The blood smears are usually observed by a person
manually and the quality of diagnosis is thus dependent on the
level training of that person. Furthermore, the time taken to
observe the smear is typically range between 15-20 minutes
[4]. However in some cases, the malaria images are blurred,
low contrast, hazy and afflicted by unwanted noises. These

978-1-61284-413-8/11/$26.00 ©2011 IEEE 257


2011 IEEE 7th International Colloquium on Signal Processing and its Applications

Figure 2.Plasmodium Falciparum in Thin Blood Smear

The research methodology involves: Where,


A. Image Acquisition
: Threshold value
The slide images for blood cell section were analysed using : Dark stretching factor
microscope at 100x magnification. Then, the images are : Colour level for the input pixel
captured and saved into bitmap (.bmp) format with resolution : Colour level for the output pixel
of 800x600. The slides have been obtained and captured at
Hospital Universiti Sains Malaysia (HUSM) under the Referring to (2), the process tends to stretch the range of
supervision of microbiologists. image value which is less than the threshold value. On the
other side, it compresses the range of image values which are
B. Image Enhancement Using Dark Stretching greater than the threshold value. Fig. 2 shows the dark
stretching process with the value of 100 is used as an example
Some major difficulties faced in image processing are
of threshold value and 150 as a dark stretching factor.
contrast and brightness problem [6]. To overcome these,
image enhancement is employed to increase the contrast of an
image. In general, an image can be enhanced by spreading the
range of colour values to make use of all possible values. The
technique is called contrast stretching. It changes the
distribution and range of the digital numbers assigned to each
pixel in an image. This is normally done to enhance image
details that may be difficult for the human viewer to observe.
Contrast stretching is one of the image enhancement
techniques that commonly be used for medical images such as
bright, dark, global and partial contrast stretching. In general
terms, contrast stretching refers to the difference in luminance
between an object and its background [7].
Dark stretching is a process that use auto scaling method Figure 3.Dark Stretching Process
which is a linear mapping function mostly used to enhance the
brightness and contrast level of the image. This technique will C. Image Segmentation
based on the original brightness and contrast level of the Image segmentation is an important technology for image
image to do the adjustment [5]. The mapping function is processing [10]. There are many algorithms used for image
shown below [5]. segmentation, and some of them segmented an image based
on the object while some can segment automatically[10]. The
goal of image segmentation is to cluster pixels into salient
image regions, i.e., regions corresponding to individual
surfaces, objects, or natural parts of objects.
Where,

: Colour level of the output pixel. III. RESULTS AND DISCUSSIONS


: Colour level of the input pixel. The proposed contrast enhancement techniques have been
: Maximum colour level values in the input image. applied to malaria slide images based on thick blood smear.
: Minimum colour level values in the input image. By implementing the dark stretching technique, the dark areas
min : Lowest pixel value in the image of the image are stretched while the bright areas are
max : Highest pixel value in the image compressed. In the malaria images, dark area is referred to the
parasites. Thus, the parasites become clearer due to the
The colour level produced is based on equation (2). stretching step in dark stretching technique.

258
2011 IEEE 7th International Colloquium on Signal Processing and its Applications
Difference controlled parameters called threshold value had [7] S.I Shahidan, Zaleha Salleh, M. Y. Mashor, A.S.W. Wahab and
H.Jaafar, “Contrast Image Enhancement in Tissue Section for
been applied on different malaria images in order to find the
Detecting Tuberculous Bacilli,” in Malaysia Technical Universities
suitable threshold value. Meanwhile the dark stretching factor Conference on Engineering and Technology (MUCET 2008) Perlis,
only used the same parameter for all images, which is 255. Malaysia, 2008.
This value is an extreme dark stretching factor that is used to [8] M. E. Al, “Region Based Contrast Enhancement of Mammograms,” in
change the background and unwanted region into white color IEEE Transaction on Medical Imaging, Volume 11, No 3, September
1992.
Here, four original images as shown in Figure 4, Figure 8 [9] N.A. Seman, N.A. Mat Isa, L.C. Li, Z. Mohamed, U. K. Ngah, K. Z.
and Figure 12 were used to test the performance of dark Zamli, “Classification of Malaria Parasite Species Based On Thin
stretching technique. The threshold value of 200 was used for Blood Smear Using Multilayer Perceptron Network,” International
Figure 5, Figure 9, and Figure 13. While the result in Figure 6, Journal of the Computer, the Internet and Management, Vol. 16. No 1
(January-April, 2008) pp 46-52.
Figure 10 and Figure 14 used 220 as threshold value and [10] Chin-Ya Huang, Mon-Ju Wu, “Image Segmentation,” ECE 533 Final
Figure 7, Figure 11 and Figure 15 used 230 as threshold value. Project, Fall 2006, University of Wisconsin- Madison.
The result shows that malaria parasite has appeared to be
clearer after the images have been process by dark stretching
technique using 220 threshold values. The only remaining is
malaria parasite. These values were found to be suitable for
the four different types of images. Based on these results, it
shows that dark stretching technique can be utilised for both
enhancement and thresholding technique on malaria slide
images.

CONCLUSION
The results show that the dark stretching technique is
potential to be used for enhancing and segmenting the malaria
slide images by adjusting only the threshold value. It should
be noticed that for the dark stretching technique, the threshold
value must be smaller than dark stretching factor. The results
in previous section show that with the proper selection of the
threshold and dark stretching factor, this technique is capable
to enhance the image quality and segment the region of
interest in malaria slide images based on thick blood smear.

ACKNOWLEDGEMENT
We would like to express our greatest gratitude to
Microbiologist Department, University Hospital, University
Science Malaysia, Kubang Kerian, Kelantan for kindly
providing valuable information on Malaria slide image.

REFERENCES
[1] World Health Organization (WHO), International
Travel and Health, 1211 Geneva 27, Switzerland, situation as on 1
January 2010.
[2] Citing internet source URL:
http://www.emedicine.com/med/byname/malaria.htm
(11 August 2010).
[3] Citing internet source URL:
http://www.netdoctor.co.uk/travel/diseases/malaria_diseases.htm
(11 August 2010).
[4] V.V. Makkapati, R. M. Rao, “Segmentation of Malaria Parasites in
Peripheral Blood Smear Image,” ICASSP 2009, March 22,2010 from
IEEE Xplore.
[5] N.A.Mat Isa, M.Y.Mashor & N.H.Othman, “Contrast Enhancement
Image Processing Technique on Segmented Pap Smear Cytology
Images”, Proc. of Int. Conf.on Robotics, Vision, Information and
Signal Processing, 118-125, 2003.
[6] E. E.Lim, P. A. Venkatachalam, U. K. Ngah, & N. E. A. Khalid, ”
Liver Disease Diagnosis by Region Growing”, Proceedings of
International Conference on Robotics, Vision and Parallel Processing
for Automation,1.38-45, 1999.

259
2011 IEEE 7th International Colloquium on Signal Processing and its Applications

(b) Ring (c) Ring (d) Ring


(a) Ring

Figure 4.Original Image Figure 5.Image after Dark Stretching Figure 6. Image after Dark Stretching Figure 7. Image after Dark Stretching
Threshold value : 200 Threshold value : 220 Threshold value : 230
Dark stretching factor : 255 Dark stretching factor : 255 Dark stretching factor : 255

(a) Ring (b) Ring (c) Ring (d) Ring

Figure 8.Original Image Figure 9.Image after Dark Stretching Figure 10.Image after Dark Stretching Figure 11.Image after Dark Stretching
Threshold value : 200 Threshold value : 220 Threshold value : 230
Dark stretching factor : 255 Dark stretching factor : 255 Dark stretching factor : 255

(a) Gametocyte (b) Gametocyte (c) Gametocyte (d) Gametocyte

Figure 12.Original Image Figure 13.Image after Dark Stretching Figure 14. Image after Dark Stretching Figure 15.Image after Dark Stretching
Threshold value : 200 Threshold value : 220 Threshold value : 230
Dark stretching factor : 255 Dark stretching factor : 255 Dark stretching factor : 255

260

Das könnte Ihnen auch gefallen