The aim of this project is to implement a classification algorithm to recognize handwritten digits (0-9).
Implementation
This project is mainly divided into three steps:
1) Machine learning model
2) Image processing
3) GUI Application
We can create this
project only by using in-built datasets of machine learning. But to visualize
the real world application we have added the image processing and application part
in this project.
Now let’s see each step
one by one in briefly.
1)
Machine learning model-:
· What is scikit-learn or
sklearn?
Scikit-learn is probably the most useful library for
machine learning in Python. The sklearn library contains a lot of efficient
tools for machine learning and statistical modelling including classification,
regression, clustering and dimensionality reduction.
Sklearn is used to build machine learning models. It should
not be used for reading the data, manipulating and summarizing it. There are
better libraries for that (e.g. NumPy, Pandas etc.)
· The digit
dataset-
This
dataset is made up of 1797 images with 8x8 pixels. Each image, like the one
shown below, is of a hand-written digit. In order to utilize an 8x8 figure like
this, we’d have to first transform it into a feature vector with length 64.
· Algorithm-
Here
we can use different classification or regression model. We have created
different classifiers/regressors and also check their accuracy . Let’s see
below table,
Classifier/Regressor
|
Accuracy
|
Logistic Regression
|
0.96
|
Stochastic Gradient Descent
|
0.95
|
K Nearest Neighbor
|
0.99
|
Linear Regression
|
0.58
|
Random Forest
|
0.84
|
As we can
see in above table, the accuracy of K nearest neighbor is more. So we have
chosen K nearest neighbor classification algorithm for this particular
handwritten digit recognition project.
K-Nearest
Neighbors:
K-Nearest
Neighbors is a Machine Learning model that tries to classify new value by
comparing it with the values of its closest neighbors.
In the example below, the arrow
points a new data point that needs to be classified into either white or black
groups. ‘K’ in K-Nearest Neighbors refers to the number of neighbors. In this
case, let's assume K=3.
So, the 3
nearest neighbors of the new data points are taken into consideration. 2 White
points and 1 black point, based on the majority we classify the new data point
as White. This is the KNN Algorithm.
·
Split the data for training and testing
sklearn.model_selection.train_test_split(*arrays, test_size=None,random_state=None)
Ø *arrays -Allowed inputs
are lists, numpy arrays, scipy-sparse matrices or pandas dataframes.
Ø test_size float or int,
default=None
If float, should be
between 0.0 and 1.0 and represent the proportion of the dataset to include in
the test split. If int, represents the absolute number of test samples. If
None, the value is set to the complement of the train size. If train
size is
also None, it will be set to 0.25.
Ø train_sizefloat or int,
default=None
If float, should be
between 0.0 and 1.0 and represent the proportion of the dataset to include in
the train split. If int, represents the absolute number of train samples. If
None, the value is automatically set to the complement of the test size.
Ø random_state
Controls the shuffling
applied to the data before applying the split.
· Code-
2) Image
Processing-:
Python
module used
1.
Image
Image
is a class of python library pillow. Pillow is a Python Imaging Library. The
Python Imaging Library adds image processing capabilities to your Python
interpreter. This library provides extensive file format support, an efficient
internal representation, and fairly powerful image processing capabilities. The
core image library is designed for fast access to data stored in a few basic
pixel formats. It should provide a solid foundation for a general image
processing tool.
Image
helps us in opening the image we get from the application and cropping it
accordingly.
2.
ImageOps
Same
as Image, ImageOps is also a class of pillow. ImageOps is used for inverting
the image and making it in greyscale.
3.
Numpy
NumPy
is a library for the Python programming language, adding support for large,
multi-dimensional arrays and matrices, along with a large collection of
high-level mathematical functions to operate on these arrays.
The
data, being large in size, must be converted to numpy arrays for faster
computations, before sending it to the machine learning models.
·
Code
3) GUI Application-:
Python modules used
1.
Tkinter
Tkinter is a Python binding
to the Tk GUI (Graphical User Interface) toolkit. It is the standard Python
interface to the Tk GUI toolkit, and is Python's de facto standard GUI. Tkinter
is included with standard Linux, Microsoft Windows and Mac OS X installs of
Python. The name Tkinter comes from Tk interface.
With the help of Tkinter,
new GUI application is created which acts as a canvas for drawing the digits.
Upon drawing digits, the app predicts the digit using image processing and
machine learning modules.
2.
tkcap
tkcap is a python project
module which takes screenshot of the tkinter window.
·
Code
Ø When the application starts,
_init_ method is started which defines the attributes of the app such as
foreground colour, background colour, pen width, etc. After this drawwidgets()
method adds widgets to the application (buttons and menubar)
Ø Among other attributes, one
attribute is model imports the knn model from
the machine learning model.
Ø When a digit is drawn on the
canvas, tkcap takes screenshot of the application and this screenshot is
processed to convert it from image format to numerical data so that it can be
utilized by knn model. Upon conversion, this data is saved for next step.
Ø Now we have the model and also the data ready
for the prediction. By clicking the prediction button we can get the
prediction. The data is fed to the knn model and it returns the predicted
value. This value is then displayed.
Experimental Result
