Colony Counter 1

8/10/2019 Colony Counter 1

1/9

International Journal of Advances in Engineering & Technology, Sept 2012.

IJAET ISSN: 2231-1963

364 Vol. 4, Issue 2, pp. 364-372

COMPUTATIONAL APPROACH TO COUNT BACTERIAL

COLONIES

Navneet Kaur Uppal1, Raman Goyal

2

1M.Tech Scholar &

2Assistant Professor,

Department of CSE, at LLRIET, Moga, Punjab, India

ABSTRACT

Research involving bacterial pathogens often requires enumeration of bacteria colonies. Bacterial

colony enumeration is an essential tool for many widely used biomedical assays. Bacterial colony enumerating

is a low throughput, time consuming and labor intensive process. Since there might exist hundreds or

thousands of colonies on a Petri dish, and the counting process is often manually performed by well-trained

technicians .There are several methods for enumeration of bacterial colonies. An increased area of focus in

Microbiology is the automation of counting methods. An increased area of focus in Microbiology is the

automation of counting methods. Several obstacles need to be addressed for methods that count colonies

present. These obstacles include: how to handle confluent growth or growth of colonies that touch or overlap

other colonies, how to identify each colony as a unit in spite of differing shapes, sizes, textures, colors, light

intensities ,etc. This method is designed to provide a degree of accuracy in counting. A colony counter is

used to count colonies of bacteria or other microorganisms growing on an agar plate. This method is used toovercome these obstacles are thresholding, segmentation, time domain frequency, Watersheding, edge

detection and morphology operator, regional descriptors etc.. This method provides high degree of accuracy.

Our proposed counter has a promising performance in terms of both precision and recall, and is flexible and

efficient in terms of labor- and time-savings.

KEYWORDS:Bacteria, Bacterial colony, Thresholding, Morphology, Watersheding.

I. INTRODUCTION

Colony counting of bacteria is crucial for quantitative, precise assessment of pathogens in clinical

research and diagnosis.Manually counting of bacteria colony is very difficult. An automated colony

counting is time saving & less labour intensive process. To provide the fast & accurate result & to

reduce the labour workload two colony counting approaches are developed. These approaches mayutilize both direct and indirect methods of counting colonies. The traditional plate count method is

considered an indirect method, desirable due to its low cost. Some examples of technologies that may

be used for quantification of microbial growth include: ATP Bioluminescence, Spiral Plating,

Membrane Filtration, Direct Epi-fluorescent Filter Microscopy, and Membrane Laser Scanning

Fluorescence Cytometry. Several obstacles need to be addressed for methods that count colonies

present. This method is designed to provide a degree of accuracy in counting that could be correlated

to the counts that would be obtained using a well-trained operator. In the automated approach ,

Bacteria are grown onto filter for 24 to 48 hours to check the contamination level of the sample. To

count these bacterial colonies microbiologist uses some dyes so that bacterial colonies appear as

colored spots and our problem is to count the number of these bacterial colonies. Further in an

Industry thousands of such samples are formed per day and colonies on each sample are counted

manually, then this becomes a time consuming hectic and error prone job. The goal of our proposed

project is to develop software to save time with accurate results and fast delivery to customers. This


2/9



365 Vol. 4, Issue 2, pp. 364-372

project will be extended in such a manner that it will count the colonies after 6 to 8 hours prior, saving

a lot more time. In this paper the proposed approaches differentiates the processing of colony images

with color information filtering is used for denoising.

II. BACTERIA

Bacteriaare unicellular microorganisms. They are typically a few micrometers long and have manyshapes including curved rods, spheres, rods, and spirals. The study of bacteria is bacteriology, a

branch of microbiology. Bacteria are ubiquitous in every habitat on earth, growing in soil, acidic hot

springs, radioactive waste, seawater, and deep in the earth's crust. A group or cluster of bacteria

derived from one common Bacterium. A bacterial cell is a unit of bacterial colony. In colony of

bacteria there can be thousands to millions of cells can be present. The morphology of the colony of

bacteria can be seen with naked eyes without the help of any gear.

A colony counter is an instrument used to count colonies of bacteria or other microorganisms growing

on an agar plate. Bacterial colony is a group of bacteria growing on a plate that is derived from one

original starting cell. An agar plate is a sterile Petri dish that contains a growth medium used to

culture microorganisms. Bacterial colony counting process is usually performed by well-trained

technicians manually.

III. METHODOLOGY

3.1. Image Capturing

Bacterial Colonies are grown onto filter for 16 to 24 hours. Some colored dyes are spread over each

filter so that bacterial colonies appear as colored spots. Now this filter is kept on a Petri plate.

Background is made of black or white intensity so, it becomes easier to separate the filter from its

surroundings while processing the image. Petri plate is kept in a box containing a digital camera and

light arrangement. Images are then captured using this arrangement. The collected images are

digitized on a computer utilizing a image processing software package that has programming

capabilities. The digitized picture is processed using the various procedures described to separate and

detect the colonies present.

3.2. Wavelet Transform

A discrete wavelet transform (DWT) is any wavelet transform for which the wavelets are discretely

sampled. As with other wavelet transforms, a key advantage it has over Fourier transforms is temporal

resolution: it captures both frequency and location information. The Haar wavelet is a sequence of

rescaled "square-shaped" functions which together form a wavelet family or basis. Wavelet analysis is

similar to Fourier analysis in that it allows a target function over an interval to be represented in terms

of an orthonormal function basis.

3.3. Extracting Image Contents

Extraction of Bacteria present on petri dish is done using various procedures. Extraction procedure

consists of following steps. Each extraction step has its own significance

3.3.1. Conversion of RGB to HSV

When it comes to computer vision, RGB values will vary a lot depending on strong or dim lighting

conditions and shadows, etc. In comparison, HSV is much better at handling lighting differences, and

it gives an easy to use color value. The advantage of HSV is that it provides a way of selecting colors

that is more intuitive to most artists.

3.3.2. Gray Scaling of Image

Grayscale digital image is an image in which the value of each pixel is a single sample, that is, it

carries only intensity information. Images of this sort, also known as blackened-white, are composed

exclusively of shades of gray, varying from black at the weakentsst intensity to white at the strongest.

Grayscale images are distinct from one-bit bi-tonal black-and-white images, which in the context of

computer imaging are images with only the two colors, black, and white (also called bi level or binary


3/9



366 Vol. 4, Issue 2, pp. 364-372

images). Grayscale images have many shades of gray in between. Grayscale images are also called

monochromatic, denoting the presence of only one (mono) color (chrome).

Fig.1.Flow Work

3.4. Applying Filter for Noise RemovalAdaptive filters adapt or learn the characteristics of the signal. Therefore the adaptive median filtering

has been applied widely as an advanced method compared with standard median filtering. The

Adaptive Median Filter performs spatial processing to determine which pixels in an image have been

affected by impulse noise. The Adaptive Median Filter classifies pixels as noise by comparing each

pixel in the image to its surrounding neighbor pixels. The size of the neighbourhood is adjustable, as

well as the threshold for the comparison. A pixel that is different from a majority of its neighbors, as

well as being not structurally aligned with those pixels to which it is similar, is labeled as impulse

noise. These noise pixels are then replaced by the median pixel value of the pixels in the

neighbourhood that have passed the noise labeling test.

3.4.1. Weiner Filter

The most important technique for removal of blur in images due to linear motion or unfocussed optics

is the Wiener filter. From a signal processing standpoint, blurring due to linear motion in a

Image capturing

Wavelet t ransform

Gray scaling

Applying filter

Thresholdin

Thinning

Eroding

Subtracting from original image

Dilation

Removing distortion

Removal of spurs

Ed e detection

Watersheding

Segregating overlapped and non overlapped

Non overlapped Overlapped

Total=overlapped +non overlapped


4/9



367 Vol. 4, Issue 2, pp. 364-372

photograph is the result of poor sampling. The inverse filtering is a restoration technique for de-

convolution, i.e., when the image is blurred by a known low-pass filter, it is possible to recover the

image by inverse filtering or generalized inverse filtering. However, inverse filtering is very sensitive

to additive noise. The approach of reducing one degradation at a time allows developing a restoration

algorithm for each type of degradation and simply combining them. The Wiener filtering executes an

optimal tradeoff between inverse filtering and noise smoothing. It removes the additive noise and

inverts the blurring simultaneously. The Wiener filtering is optimal in terms of the mean square error.

In other words, it minimizes the overall mean square error in the process of inverse filtering and noise

smoothing.

1 2 1 21 2 2

1 2 1 2 1 2

( , ) ( , )( , ) ,

( , ) ( , ) ( , )

xx

xx

H f f S f fW f f

H f f S f f S f f

=

+

Where, Sxx(f1,f2), S(f1,f2) are respectively power spectra of the original image and the additive

noise, and H (f1,f2) is the blurring filter.

3.5. Thresholding

It is one of the methods of image segmentation. This method is based on a clip-level (or a threshold

value) to turn a gray-scale image into a binary image. Thresholding can be defined as mapping of the

gray scale into the binary set {0, 1} that is thresholding essentially involves turning a color or

grayscale image into a 1-bit binary image. Individual pixels in a gray scale image are marked as

object pixels if their value is greater than some threshold value and as background pixels

otherwise. Generally the object pixels are black and the background is white. After thresholding a

binary image is formed where an object pixel is given a value of 1 while a background pixel is given

a value of 0. The key of this method is to select the threshold value (or values when multiple-levels

are selected). Several popular methods are used in industry including the maximum entropy method,

Otsu's method (maximum variance), and et al. k-means clustering can also be used. & the main

purpose of thresholding is to discard irrelevant data and keep only the important segments of data

which lie above threshold curve to Classify pixels on the basis of color or a local property. To providea bi-level picture that distinguishes objects from background. To represent the text and diagrams in

binary formats.

3.6. Thinning

It is a morphological operation that is used to remove selected foreground pixels from binary images,

somewhat like erosion or opening. It can be used for several applications, but is particularly useful for

skeletonization. In this mode it is commonly used to tidy up the output of edge detectors by reducing

all lines to single pixel thickness. Thinning is normally only applied to binary images, and produces

another binary image as output.

3.7. MorphologyMorphological image processing (or morphology) describes a range of image processing techniques

that deal with the shape of features in an image..The language of mathematical morphology is set

theory. The mathematical morphology simplifies image data, preserves essential shape characteristics

and eliminates noise. Morphological operations are based on simple expanding and shrinking

operations. The primary application of mathematical morphology occurs in binary images, though it is

also used in grey scale images. Morphological operations are typically applied to remove

imperfections introduced during segmentation. Mathematical morphology is a foundation of

morphological image processing, which consist of a set of operators that transform images according

to the above characterizations. The morphological operator dilation acts like a local maximum

operator. Dilation generally increases the sizes of objects, filling in holes and broken areas, and

connecting areas that are separated by spaces smaller than the size of the structuring element. Erosion

acts like a local minimum operator. Erosion generally decreases the sizes of objects and removes

small anomalies by subtracting objects with a radius smaller than the structuring element.


5/9



368 Vol. 4, Issue 2, pp. 364-372

3.8. Edge DetectorIt is a process that detects the presence and location of edges constituted by sharp changes in color

intensity of an image. . For a good edge detection, the edge line should be thin and with few speckles.

The purpose of edge detection in general is to significantly reduce the amount of data in an image,

while preserving the structural properties to be used for further image processing.Cannys aim was to discover the optimal edge detection algorithm. Sobel and Prewitt which uses first

derivative have very simple calculation to detect the edges and their orientations but have inaccurate

detection sensitivity in case of noise.

3.9. WatershedingThis is method of image processing segmentation. There is a need to split the colony to get correct

colony counts. To separate the connected colonies intensity are considered. Intensity gradient image

as a topological surface, thus Watershed algorithm can be applied to divide the clustered colonies in

the image just as water fold in the topological surface. Think of the grey-level image as a landscape.

Let water rise from the bottom of each valley (the water from each valley is given its own label). As

soon as the water from two valleys meets, build a dam, or watershed. These watersheds will then

define the borders between different regions. Watershed can be used directly on the image, on an edgeenhanced image or on a distance transformed image. A grey-level image may be seen as a topographic

relief, where the grey level of a pixel is interpreted as its altitude in the relief. A drop of water falling

on a topographic relief flows along a path to finally reach a local minimum. Intuitively, the watershed

of a relief corresponds to the limits of the adjacent catchment basins of the drops of water The

gradient magnitude of an image is considered as a topographic surface for the watershed

transformation. Watershed lines can be found by different ways. The complete division of the image

through watershed transformation relies mostly on a good estimation of image gradients. The result of

the watershed transform is de-graded by the background noise and produces the over-segmentation.

Also, under segmentation is produced by low-contrast edges generate small magnitude gradients,

causing distinct regions to be erroneously merged.

Our proposed method can recognize chromatic and achromatic images and thus can deal with both

color and clear medium. In addition, our proposed method can also accept general digital cameraimages as its input. The whole process includes detecting dish/plate regions, identifying colonies,

separating aggregated colonies, and finally reporting consistent and accurate counting results.

IV.

RESULTS

TEST 1(a)(Processed image of Non Overlapped Bacterial Colonies)

Figure 1.1 Processed image of Non Overlapped Bacterial Colonies

Figure 1.2 Total number of Overlapped Bacterial Colonies found


6/9



369 Vol. 4, Issue 2, pp. 364-372

TEST 2(a)(Processed image of Non Overlapped Bacterial Colonies)

Fig. 2.1:Processed image of Overlapped Bacterial Colonies

Fig. 2.2:Total number of Overlapped Bacterial Colonies found

Table 1.Result Image 2. jpg

Sr No. Fig Name Overlapped Total

1 2.jpg 20 100

TEST 2(b)(Processed image of Overlapped Bacterial Colonies)



TEST 2(c)(Processed image of Overlapped Bacterial Colonies)


7/9



370 Vol. 4, Issue 2, pp. 364-372



TEST 2(d)(Processed image of Overlapped Bacterial Colonies)



Table 2. Manually Count & Proposed Method Count

SR.NO. FIG.

NAME

TOTAL

COUNT

MANNUALY

OVERLAPPED NON-

OVERLAPPED

PROPOSED

COUNT

METHOD

1 Fig.1.1 64 0 64 64

2 Fig.2.1 102 20 80 100


8/9



371 Vol. 4, Issue 2, pp. 364-372

3 Fig.3.1 10 2 8 10

4 Fig.4.1 17 3 14 17

5 Fig.5.1 59 11 46 57

V.

CONCLUSION

From the results and comparison of the different methods of edge detection, it is concluded that the

Computational Bacterial Colony Counter is better than the traditional method of manual

counting of bacterial colonies on petri dish.

Figure 6.Proposed count and Manual count

We can design an automated bacterial colony counter which used many image processing algorithms

such as grayscaling, thresholding, filtering, Watersheding etc. to count these colonies efficiently.Hence, Images with high contrast and low or medium density give accurate or near to accurate count

(99-100%). Images with low contrast or high density give less accurate count (95-98%) as given in

graph. The reason thereof is that in case of low contrast after thresholding shape of colony/colonies

gets distorted which leads to appearance of high curvature points along the boundary. These high

curvature points (corners) get accumulated in count result.

VI.

FUTURE SCOPE

Image processing is very advanced and extensive field which needs extensive research and hard work.

1. This work can be extended to run on different types of micro biological bodies in order to

count and evaluate their different parameters like shape, density etc..2. Parallel Computing can also be done in order to speed up the process.

3. Process the most complex samples and give accurate count.

REFERENCES

[1] Mathew L. Clarke, Robert L. Burton, A. Nayo Hill, Maritoni Litorja, Moon H. Nahm, Jeeseong Hwang,

Low cost, High throughput, automated counting of bacterial colonies, Journal of the International Society

for Analytical cytology vol. 77, issue 8, pp 790-797, [2010] .

[2] SHEN Wei-zheng, WU Ya-chun, ZHAO lie, ZHENG Hui, Experimental Study for Automatic Colony

Counting System Based on Image Processing, 2010 International Conference on Computer Application

and System Modeling [ICCASM 2010].


9/9



372 Vol. 4, Issue 2, pp. 364-372

[3] S.Lakshmi, Dr. V. Sankaranarayanan, A study of Edge Detection Techniques for Segmentation Computing

Approaches, IJCA Special Issue on Computer Aided Soft Computing Techniques for Imaging and

Biomedical Applications CASCT, 2010. pp35-41

[4]

Rafael C. Gonzalez and Richard E. Woods, Digital Image Processing3rd Edition, Addison Wesley, 2009

[5]

Chengcui Zhang, Wei-Bang Chen, Wein-Lin Liu and Chi-Bang Chen. An automated Bacterial Colony

counting system, IEEE International Conference on Sensor Networks, Ubiquitous and TrustworthyComputing, pp 233-240, [2008].

[6]

Ehsan Nadernejad, Edge detection techniques : Evaluation & Comparision, Applied Mathematical

Sciences vol.2,no.31,1507-1520[2008]

[7]

Bang Wei, Zhang and chengui Chen. An effective and robust method for automatic bacterial colony

enumeration, proceeding of the international workshop on semantic computing and multimedia system in

conjunction with international conference on semantic computing vol. 17,pp 581-588,[2007].

[8]

Mohammed Roushdy,Comparative study of edge detection algorithms applying on the greyscale noisy

image using morphological filterGVIP,Journal ,volume 6,issue 4,December,[2006].

[9]

Michael Putman, Robent Burton, Moon H. Nahm. Simplified method to automatically count bacterial

colony forming unit, Journal of Immunological Methods, pp 99-102, [2005].

[10] Martin, D., Fowlkes , C., Malik, J.: Learning to detect natural image boundariesusing brightness and

texture. IEEE Transactions on Pattern Analysis and Machine Intelligence 26 (5) pp. 530549 [2004].

[11]

S.K.park,b.k.choi, Boundary extraction algorithm for cutting edge detection,computer aided design no. 33

pp.571-579 [2001].

[12]

Ma, W.Y., Manjunath, B.S.: Edgeflow: A technique for boundary detection andsegmentation. IEEE

Transactions on Image Processing 9 (8) pp 13751388 [2000].

[13]

Kawai M ,yamaguchin and Nasu M. Rapid enumeration of Physiologically active bacterial in purified water

used in pharmaceutical manufacturing process, Journal of applied microbiology, vol.86, pp496-

504.[1999].

[14]

Paul.C.Olsztyu, James H.Beyar, David Sullivan. Bacterial colony counter and classifier, vol.83, pp382-

400, [1998].

[15]

S.M. Smith and M. Brady. SUSAN - A New Approach to Low Level Image Processing, International

Journal of Computer Vision, Vol. 23(1), pp. 45-78, [1997]

[16]

E.Yourassowsky and E.Schoutens.An automated count and size evaluation of colonies of bacteria grown

in a zonal concentration gradient of antimicrobial agent, American society for microbiology, vol.28 no.4,

pp 525-529 [1994].

[17]

A. Marion An Introduction to Image Processing, Chapman and Hall, pp. 274.[1991].

[18]

J. Chen, A. K. Jain, "A Structural Approach to Identify Defects on Textural Images", Proceedings of the

IEEE International Conference on Systems, Man, and Cybernetics, pp. 29-32, Beijing, [1988].

[19]

Nabil Jean Naccache and Rajan Shinghal, Member IEET, SPTA: A proposed thinning algorithm for binary

digital patterns, IEEE transaction on system, man, cybernetics, vol. S.N.C 14, No. 3. [1984].

[20]John M.Kramer, Margaret Kendall and Richard J.Gilbert. Evaluation of the spiral plate and Laser Colony

Counting techniques for the enumeration of bacteria in foods, journal of applied Microbiology and

Biotechnology,Vol.6,pp.289-299[ 1978].

AUTHORS BIOGRAPHY

Navneet Kaur Uppal was born in Punjab, India, in 1988. She received her Bachelor degree

in Computer Science Engg. degree from Amritsar College of Engineering and Technology

in 2009 and persuing the Masters from the Lala Lajpat Rai Institute of Engineering and

Technology .

Documents

Colony Counter 1