Flower Identification App using Machine Learning

Project Title

Flower Identification App using Machine Learning

Project Area of Specialization Artificial IntelligenceProject Summary

The Flower is the beauty of nature. We see different flowers while strolling outdoor in a park, garden, roadside, farm, or while traveling. A naive person usually doesn’t have knowledge about all of them and sometimes doesn’t even know their name. What if we see a beautiful flower and wonder what kind of flower is it? Although we can consult flower guidebooks or browse over internet but it’s a bit difficult to find the name and some flowers are similar in look. We propose a mobile app, that will recognize some classes of flowers by using Tensorflow (an Open-source Dataflow, Python Library for Machine Learning) on Android platform. We will use Google Colab for Training Flower Model.

Project Objectives

The main objective of our project is to make it easy to recognize/identify some classes of flowers using a Machine Learning Model that will classify flower on the basis of its features rather than assumptions.

Project Implementation Method

• Collecting Flower Dataset
• Install required packages and Import Libraries

• Data Loading and Data Splitting

• Train Model from split data
• Evaluate the Model from test Data.
• Export to TensorFlow Lite model.
• Import model to android studio and Implementation
• Documentation

Data Set that will be used for training and testing is obtained from Tensor Flow dataset repository [1]. The flowers dataset consists of 5 classes of flowers, each class having more than 600 images. When training a machine learning model, we will split our data into training and test datasets. We will train the model on our training data and then evaluate by testing data.

After that the trained model will be implemented in android Application using tensor flow lite library.

Benefits of the Project

This project will provide help to researchers and developers for further work and improvement. It will also be helpful for Scientists or Botanists to classify large amounts of data and generate the required results with further improvement. In addition, it would be really helpful while spending time outdoors to identify flowers by a mobile application. Moreover, it would be very beneficial for a Traveler, Hiker, or Nature Enthusiast to easily identify flowers.

Technical Details of Final Deliverable

• Collecting Flower Dataset
• Install required packages and Import Libraries

• Data Loading and Data Splitting

• Train Model from split data
• Evaluate the Model from test Data.
• Export to TensorFlow Lite model.
• Import model to android studio and Implementation

Final Deliverable Android Application.

Final Deliverable of the Project Software SystemCore Industry ITOther Industries Agriculture Core Technology Artificial Intelligence(AI)Other Technologies OthersSustainable Development Goals Life on LandRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	9900
Internet	Miscellaneous	12	700	8400
Stationery	Miscellaneous	3	500	1500