You are on page 1of 16

G. H.

Raisoni College of Engineering, Nagpur


Department of Information Technology

Document Image Augmentation


  Department of Information Technology

Presented by:

Name Roll No
1)Vikas singh 72
2)Yuvaraj Dakhane 73
3)Surbhi Fule 74
4)Tejal Kale 75
5) Himamshu Doye. 30

Project Guide: Prof. Nirmal Mugale

.
CONTENTS

Problem
 statement?
Why we need of large amount of data?

Aim

What is need of Image Augmentation?

Objective

Features

Flowchart

Technology Used

Benefits of image augmentation

Conclusion

Future Scope

Problem Statement?

While considering deep learning methods to train


our data we often come across two problems, it
requires — huge amount of training data and
powerful computational infrastructure.
Why we Require large amount of data?

When you train a deep learning model, what you’re really doing is tuning its
parameters such that it can map a particular input (e.g. an image) to some
output (a label). Our optimization goal is to chase that sweet spot where our
model’s loss is low, which happens when your parameters are tuned in the
right way. Naturally, if you have a lot of parameters, you would need to
show your machine learning model a proportional amount of examples, to
get good performance. Also, the number of parameters you need is
proportional to the complexity of the task your model has to perform.
Aim

Document Image Augmentation is simple to use


tool for performing Augmentation
Document image analysis and understanding is to
automatically recognize and extract textual or
graphical material from digitized documents.
Document Image Analysis addresses the problem
of separation of text and graphics and their
identification and recognition.
WHAT IS NEeD of Image Or Data
Augmentation ?
• Data augmentation is a strategy that enables practitioners to
significantly increase the diversity of data available for training
models, without actually collecting new data. Data augmentation
techniques such as cropping, padding, and horizontal flipping are
commonly used to train large neural networks

• Image data augmentation is a technique that can be used to


artificially expand the size of a training dataset by creating
modified versions of images in the dataset. ... Image data
augmentation is used to expand the training dataset in order to
improve the performance and ability of the model to generalize.
Data augmentation is an integral process in deep
learning, as in deep learning we need large amounts of
data and in some cases it is not feasible to collect
thousands or millions of images, so data augmentation
comes to the rescue.
• It helps us to increase the size of the dataset and
introduce variability in the dataset.
Data augmentation is a basic technique to increase a
dataset without new data. Although the technique can
be applied in a variety of domains, it’s very commonly
used in computer vision, and this will be the focus of
the post.
Objective

The objective of document image is to recognize the text


and graphics components in images, and to extract the
intended information as a human would. Two categories of
document image analysis can be defined.
It is a technique that can be used to artificially expand the
size of a training dataset by creating modified versions of
images.
It is used to artificially expand the size of a training dataset
in order to improve the performance and ability of the model
to generalize.
FEATUREs:-

This tool provide following document


Augmentation:-
Dilate document image
Smudge document image
Add support for Ground Truth
Add Flip and Rotate
Add Another Document Specific
Augmentation
Modify Color Changer to automatically
detect background color
Flowchart
Technology Used

Python
Requirements: Numpy, Opencv-
python and keras
* install numpy
* install opencv-python
* install keras
The benefits of Image Augmentation

The first is the ability to generate ‘more data’ from


limited data
The second one is to avoid overfitting: For a network it
is somewhat problematic to memorize a larger amount
of data, as it is very important to avoid overfitting. This
occurs because the model memorizes the full dataset
instead of only learning the main concepts underlying
the problem. To summarize, if our model is overfitting,
it will not know how to generalize and, therefore, will
be less efficient.
Higher accuracy is achieved if you have a larger dataset
of images to train.
FUTURE SCOPE

Data augmentation has been shown to produce promising ways


to increase the accuracy of classification tasks.
• We experimented with our own way of combining training
images allowing a neural net to learn augmentations that best
improve the ability to correctly classify images.
• If given more time, we would like to explore more complex
architecture and more varied datasets. To mimic industrial
applications, using a VGG16 instead of Small Net can help
us determine if augmentation techniques are still helpful
given complex enough networks that already deal with many
over fitting and regularization problems.
Conclusion

•The effectiveness of data augmentation through


simple techniques, such as cropping, rotating, and
flipping input images. We artificially constrain our
access to data to a small subset of the Image Net
dataset, and compare each data augmentation
technique in turn.
Thanks

You might also like