Digital Image Processing Techniques for Detecting Plant or Animal Diseases

Studienarbeit, 2015
12 Seiten



To research on the image of proliferating epithelium is a sheet of dividing cells that adhere to each other strongly. Our research how these sheets divide is important both for development and disease diagnosis. The arrangement of cell division is one of the key mechanisms that drive the development of living organisms. Abnormal divisions lead to cancer. Our research focuses on cellular shape characteristics. In this research intends to propose a topological framework that yields new insights into the underlying mechanisms that govern cell shape dynamics in proliferating epithelia. A recent study strongly suggests that cell division planes are not chosen at random, but is instead actively regulated to suppress variation in cell shape within a proliferating epithelium. To prove the theory, this study intends to identify epithelial cells which were counted automatically by a code written in Matlab 2013b. Then automatic segmentation and morphometric analysis of cell cytoplasm and nucleus was performed. Finally geometric features of normal oral epithelial cells like cellular and nuclear area and perimeter, cellular and nuclear form perimeter and contour index have been studied and presented. These data will in turn help in epithelial topology assessment .The differences in the normal and abnormal images for various diagnostic variables will be compared then it will give us better insight on computer assisted automated disease diagnostics.

1. Introduction

The epithelium Word meant only the skin on the breast. Epithelia can occur as sheets of cells, as in the lining of the intestine, or as solid aggregations of cells, as in glandular organs1. A structural distinction is made between the first type, i.e., covering and lining epithelium and second type referred to as secretory or glandular epithelium1. It found where gas exchange occurs, in small intestine, kidney, pancreas, oesophagus, salivary glands, female reproductive organs, embryo and The skin is made of epithelial cells1 -3.

Regulation of cell growth and division in plant and animal plays fundamental roles for organ development and cancer progression8 -13. Proliferating epithelial sheet of dividing cells that adhere to each other tightly, is a model used to study growth and division of cell14,15.To study growth and division of cell, we should know cell geometry and topology (topology means study of geometrical properties of an object) such that cell geometry refers cell shape, size, length, angle and boundaries14 -16 where as cell topology refers cell connectivity and no of cell neighbours14,17. Cell geometry and topology are tightly connected18, example: cell sides are linearly correlated with cell size16 and cell area dependent on growth rate19. Cell topology modified by changes in cell-cell contacts, which occur in biological processes, division, rearrangement and death of cells20 -23. Geometrical properties of cells give mechanisms of regulating cell topology in proliferating epithelia24. The topological structure of proliferating epithelia has been studied since 20th century18,25 -27.

Its functions are: no. 1- Protection: These cells replace damaged or dead cells. No.2-Sensory: It provides signals for sensory sensations e.g. taste, sight and smell. No.3-Transportation: Found on the intestinal lining and transport the filtered material such as glucose. No.4-Absorption: It absorbs the filtered material, such as glucose. No.5-Secretion: some epithelial cells, like Goblet cells, secrete fluids that are necessary for digestion, protection etc. No.6-Movement: Some epithelial cells have cilia, which aid in moving substances in the lumen2 -7. Covering and lining epithelium is classified on the basis of two criteria: The number of cell layers present and the shape of the cells in the top layer. If an epithelial sheet is composed of only one layer it is a simple epithelium. If there is more than one layer, then it is a stratified epithelium. The features of epithelia have capacity for regeneration. In most epithelia, production of new cells is a more or less constant process, and one can express the activity of a tissue in terms of it mitotic index, the ratio of dividing cells to non-dividing ones. Example: the mitotic activity of epithelia is in the small intestine. It has dense irregular shapes where the nuclei would normally be located and second example, you certainly wouldn't expect a high mitotic index in a normal salivary gland or in the kidney, though these are both epithelial organs2 -7. A high MI (mitotic index) in the intestine is an adaptation to rapid loss of cells; in other places it's a sign of uncontrolled cell division and one of the hallmarks of cancer1.In this research we examine on 853 cytoplasm31 to read their shape and size for all stages of disease from normal to abnormal condition and we also find the rate of growth and at last we plot Graphs on the basis of results which were came after applied our algorithm.

2. Experimental section

Objective :

1. Automatization for computer based cell counting in oral epithelial tissue histogy image.
2. Computer based cellular geometric feature assessment of cellular feature like cytoplasmic and nuclear area, perimeter, form perimeter and contour index.


I took images31 of epithelial cell and then applied the algorithm by the help of image processing for normal and abnormal cells of different stages.

illustration not visible in this excerpt

The image’s types are: no.1- Binary Image : An image containing only black and white pixels and represented by a unit 8 or double logical matrix containing 0’s and 1’s. No.2- Index Image: An Index image is represented by an array of class unit 8, unit 16 or double. The color map is always an m-by-3 array of class double. No.3- Intensity Image: Consisting of gray scale values. No.4- RGBv Image: Each pixel is specified by three values-red, green, blue. It represented by an m-by-n-by-3 array of class unit 8, unit 16, double 32. In our research we took the RGBv image (fig. 1)31

illustration not visible in this excerpt

fig. 1

illustration not visible in this excerpt

fig. 2

illustration not visible in this excerpt

fig. 3

and it has dimension 256, now in our algorithm we used the image processing (Matlab 2013b) , first we find the background of the image (fig. 2) then we changed the image to 2-dimension (fig. 3) , now we filtered to get sharp image. Ones we have got filtered image then we can count total no of objects (cytoplasm) present in the image by complement of the image and then find boundaries of the objects (fig. 4),

illustration not visible in this excerpt

fig. 4

illustration not visible in this excerpt

fig. 5

illustration not visible in this excerpt


now to cover the boundaries of nucleus take the filtered image(fig. 3) then Convert RGBv to YCbCr color space and apply threshold (fig. 5) to get segmentation mask now use k-mean clustering (fig. 6). The filtered image is a fairly low-contrast image, so we have used an automatic method such as adapthisteq to increase contrast then apply a threshold (fig. 5) and clean that up and then overlay the perimeter on the original image. It will cover the perimeter of each cell in the epithelial sheet (fig.7, 8), now it is easy to select (fig.9) each cytoplasm and nucleus for ROI manager in ImageJ (open source image processing) to find the Area (µm2), Perimeter (µm), Standard deviation etc.

illustration not visible in this excerpt

fig. 7

illustration not visible in this excerpt

fig. 8

illustration not visible in this excerpt

fig. 9

YCbCr also written as YCBCR or Y’CBCR is a family of color space used as a part of the color image pipeline in video and digital photography systems, y’ is the luma component and CB and CR are the blue-difference and red-difference chroma components, Y which is luminance, meaning that light intensity is non linearly encoded based on gamma corrected RGBv primaries28.


Ende der Leseprobe aus 12 Seiten


Digital Image Processing Techniques for Detecting Plant or Animal Diseases
ISBN (eBook)
ISBN (Buch)
1150 KB
digital, image, processing, techniques, detecting, plant, animal, diseases
Arbeit zitieren
Yogash Joshi (Autor)Debjani Chakraborty (Autor)Jyotirmoy Chatterjee (Autor), 2015, Digital Image Processing Techniques for Detecting Plant or Animal Diseases, München, GRIN Verlag,


  • Noch keine Kommentare.
Im eBook lesen
Titel: Digital Image Processing Techniques for Detecting Plant or Animal Diseases

Ihre Arbeit hochladen

Ihre Hausarbeit / Abschlussarbeit:

- Publikation als eBook und Buch
- Hohes Honorar auf die Verkäufe
- Für Sie komplett kostenlos – mit ISBN
- Es dauert nur 5 Minuten
- Jede Arbeit findet Leser

Kostenlos Autor werden