An image preprocessing and augmentation library with Keras like interface.

Table of contents:

• Table of contents:
• Introduction
• The Keas interface
• Install
• Dependencies
• Documentation
• Example
• Blogs and Tutorials
• Support
• Contributors
• License

Introduction

Hocrox is an image preprocessing and augmentation library. It provides a Keras like simple interface to make preprocessing and augmentation pipelines. Hocrox internally uses OpenCV to perform the operations on images. OpenCV is one of the most popular Computer Vision library.

Here are some of the highlights of Hocrox:

• Provides an easy interface that is suitable for radio pipeline development
• It internally uses OpenCV
• Highly configurable with support for custom layers

The Keas interface

Keras is one of the most popular Deep Learning library. Keras provides a very simple yet powerful interface that can be used to develop start-of-the-art Deep Learning models.

Check the code below. This is a simple Keras code to make a simple neural network.

model = keras.Sequential()
model.add(layers.Dense(2, activation="relu"))
model.add(layers.Dense(3, activation="relu"))
model.add(layers.Dense(4))

In Hocrox, the interface for making pipelines is very much similar. So anyone can make complex pipelines with few lines of code.

Install

To install Hocrox, run the following command.

pip install Hocrox

Dependencies

Hocrox uses OpenCV internally so install it before.

Documentation

Documentation for Hocrox is available here.

Example

Here is one simple pipeline for preprocessing images.

from hocrox.model import Model
from hocrox.layer import Read, Save
from hocrox.layer.preprocessing.transformation import Resize
from hocrox.layer.augmentation.flip import RandomFlip
from hocrox.layer.augmentation.transformation import RandomRotate

# Initalizing the model
model = Model()

# Reading the images
model.add(Read(path="./images", name="Read images"))

# Resizing the images
model.add(Resize((224, 244), interpolation="INTER_LINEAR", name="Resize images"))

# Augmentating the images
model.add(
    RandomRotate(
        start_angle=-10.0, end_angle=10.0, probability=0.7, number_of_outputs=5, name="Randomly rotates the image"
    )
)
model.add(RandomFlip(probability=0.7, name="Randomly flips the image"))

# Saving the images
model.add(Save("./preprocessed_images", format="npy", name="Save the image"))

# Generating the model summary
print(model.summary())

# Transforming the images
model.transform()

Blogs and Tutorials

Check this video to learn more about Hocrox.

Support

If you are facing any issues using Hocrox, then please raise an issue on GitHub or post something on the discussion section.

Alternatively, you can send email to.

Contributors

Check the list of contributors here.

License

MIT