This document summarizes a project that uses image processing to extract red blood cells from blood smear microscope images and count the cells. The process involves preprocessing the images using techniques like histogram equalization, contrast adjustment, and morphological operations. Individual red blood cells are then extracted and classified using neural networks. The overall method was able to separate and count red blood cells with 80% accuracy.

1. RED BLOOD CELLS
EXTRACTION AND COUNTING
GUIDED BY:
MR.N.KIRUBAKARAN
HOD (CSE DEPT)
PRESENT BY:
Rahul Reghunath
By using software

2. INTRODUCTION
•Blood is a connective tissue consisting of cells
suspended in plasma.
•The most abundant small reddish cells are
erythrocytes and called red blood cell.
•The conventional device used to count blood cells is
the hemocytometer
•Several attempts have been made to mimic the
procedure of cell recognition from image like
problems will come in the conventional
method.Here will our project application will come.

3. EXPERIMENT
• This work aims to apply image processing to extract the blood image
taken from blood smear microscope, then automatically counting red
blood cells .
• Digital image processing was extensively used in this work. It is the
key performance index to establish the ability of the proposed
method.
• The experiment is going through different steps.

4. STEPS
Image processing
• The main image processing tasks consists of enhancing the image's
qualities and deleting overlapped blood cells in the boundary area of
the image
Histogram equalization
• This process adjusts intensity values of the image by performing
histogram equalization involving intensity transformation.

5. Red Blood Cell counting procedure
Image processing
Single blood cell extraction
Single cell analysis and classification by
Neural Network
Red blood cells counting

6. • To adjust brightness of an image, an histogram of the
interested image is used to determine data and display
ranges of the image.
Cell detection
•The objective of blood cell detection is to detect
cells which differentiate themselves from the
background in terms of contrast
Contrast and brightness adjustment

7. Image dilation
• The dilation morphological operator has been used to better connect
separated points of the membrane.
Interior gap filling
• Filling internal holds of the connected element get the biggest area in
the processed image
Object smoothening (Erosion)
• This step reduces the spur elements along the membrane edges.

8. Single blood cell extraction
• This method extracts the single blood cell from the derived binary
image to obtain cell’s position.
Border padding
• The missing pixels will be padded using 0 value (black) to complete
the image.
Centroid finding
• The centroid of the converted binary image is measured by finding
the center of mass of the binary image region.

9. Transferringoriginal RGB image to grey
Original image.
Step 1. Equalizing image,
Step 2 .Adjusting of an
Image.
Step 3. Detecting entire
cell
Step 4. Dilating an image
Step 5. Filling interior gaps
Step 6. Smoothening an object.
(Erosion)

10. CONCLUSION
‣This worked to study the possibility of RBC using image processing.
‣ The single blood cell extracted and finally seaperated RBC offers 80%
of accuracy or better.
‣ Higher accuracy increased when the number of sample training
images is increased.