You are on page 1of 2

Description of design and implementation

Segmentation
Segmentation is one of the crucial steps in building the patten recognition system. It is
a process of broken down or partitioning the digital images into various subgroups,
which known as image segments that assist in decreasing the complexity of images.

In this project, we had manually cropped out the valuable portion of the hotels from
the background of the images. The name along with the logo were segmented together
for each of the hotels. For the example,

Feature extraction
Feature extraction is a method of reducing data dimensionality by encoding related
information in a compressed representation and removing less discriminative data.
Feature extraction is vital to gesture recognition performance. Therefore, the selection of
which features to deal with and the extraction method are probably the most significant
design decisions in hand motion and gesture recognition development. Here we used
principal component as main features.

LDA is a statistical technique that is used for dimensionality reduction in a data set. It
is also known as Latent Dirichlet Allocation. LDA is an unsupervised algorithm,
which means that it does not require any training data.LDA is used to find these latent
variables so that they can be used to improve the predictive power of machine
learning models. For instance, if you have a data set with many features (columns),
you can use LDA to reduce the dimensionality of the data set and make it easier to
build a machine learning model. LDA is also often used as a preprocessing step for
other machine learning algorithms such as support vector machines (SVMs) and
logistic regression.

We had transformed the images of the training (X_train) and testing set
(X_test) with through LDA with the objective to de-noise the data, reduce the data
dimensionality and speed up the training and testing computation time.

Classification
The classification algorithms that choose to perform this Bank Pattern
Recognition is Support Vector machines, abbreviated as SVM. SVM is one of the
supervised machine learning approaches that use for binary classification or
regression problem statements. It performs the sophisticated data transformation
based on the selected kernel. The goal is to discover the hyperplane that maximize the
separation boundaries between the data points according to the labels.

Besides that,we also chose the K-Nearest Neighbors (KNN) algorithm, chosen
for this study,it is a well-known supervised learning method for classification tasks. It
is particularly suitable for image classification because it can handle complex decision
boundaries and capture patterns and similarities between images without assuming
any particular data distribution.In KNN, the classification of a new data point is
determined by considering the class labels of its nearest neighbors in the feature
space. The algorithm stores the feature vectors and the corresponding class labels of
the training dataset. In the prediction phase, it calculates the distance between the test
samples and all training samples using distance metrics such as Euclidean distance.
The K nearest neighbors with the smallest distance are then selected.

You might also like