You are on page 1of 1

Preprocessing:

1] Pixel segmentation.

2] Pixel range decide 300 to 600.

3] Normalization to into bits 0 and 1.

Feature Extraction from Image.


-> cell boundaries, cell shape and background and cell difference.

Classification Of CNN.
-->Healthy cell and sickle cell and irrregular shape.

--> Accuracy about 90%.

--> Datasets are available in png and jpg CNN is works only in that is used in
machine learning for more accuracy.

Convolution Neural Networks.


--> CNN take image input and give output to watershed.
--> Watershed algorithm differeciate the cell boundary the and background.
--> Watershed divide it into different segments.
--> and then It output gives again to CNN.

After that the we use randam split 80% Training and 20% Testing this is dataset.

CNN have 3 layers Input layer, Hidden layer and Output layer.

Python 3.12
Spyder 5.4.3
Anaconda Navigator 2.4.2

Python Libraries:
1] Tensorflow --> It used to create the dataflow graphs.
2] Matplot lib --> It is used to draw points and lines in diagram.
3] Keras --> It is used to make the implementation of neural network easy.
4] Pillow --> It is used to show the image functions inbuilt in it like load image
and create new image.
5] Numpy --> It is used to perform mathematical operations on arrays.
6] Opencv-python --> It is used to allows you perfom image processing and computer
vision tasks.
7] tkvideo --> It is module in python used to playing video in GUIs.
8] gtts --> Google Text to Speech is used to convert text to speech in python.
9] Pandas --> It used for working with datasets.
10] Flask --> Creating a web applications in the python language.

You might also like