Repository files navigation

This project aims to classify chicken fecal samples into two categories: diseased (Coccidiosis) and healthy. The classification is based on analyzing images of the fecal samples using computer vision techniques.

The project follows a modular structure, consisting of several stages and pipelines which includes :-

1. stage_01_data_ingestion.py: This stage is responsible for data ingestion. It includes functions for downloading, extracting, and preprocessing the dataset.
2. stage_02_prepare_base_model.py: In this stage, the base model for the classification task is prepared. It involves loading a pre-trained model, modifying it if necessary, and preparing it for training.
3. stage_03_training.py: The training stage is responsible for training the model using the prepared dataset. It includes functions for data augmentation, model training, and saving the trained model.
4. stage_04_evaluation.py: This stage focuses on evaluating the performance of the trained model. It includes functions for loading the trained model, performing inference on test data, and calculating evaluation metrics.

To run this project, you need the following dependencies:

• Python (version 3.8 or above)
• TensorFlow
• Flask
• DVC

Make sure you have installed the required dependencies before running the project.

• Update config.yaml
• Update secrets.yaml [Optional]
• Update params.yaml
• Update the entity
• Update the configuration manager in src config
• Update the components
• Update the pipeline
• Update the main.py
• Update the dvc.yaml

1. Clone the repository:

git clone https://github.com/Pratik94229/Chicken-Disease-Classification-Project

cd Chicken-Disease-Classification-Project

2. Create a conda environment after opening the repository

conda create -p venv python==3.8 conda activate venv/

3. Install the dependencies:

pip install -r requirements.txt

4. Finally run the following command

python app.py

5. Open Terminal

dvc initdvc repro

6. For visualizing pipeline

dvc dag

7. Flask App:

To create a front-end interface for the application, run the Flask app:

python app.py

8. Access the app:

Open your browser and go to

http://localhost:5000 to access the application.

#with specific access1. EC2 access : It is virtual machine2. ECR: Elastic Container registry to save your docker image in aws#Description: About the deployment1. Build docker image of the source code2. Push your docker image to ECR3. Launch Your EC2 4. Pull Your image from ECR in EC25. Lauch your docker image in EC2#Policy:1. AmazonEC2ContainerRegistryFullAccess2. AmazonEC2FullAccess

- Save the URI: 738400679807.dkr.ecr.ap-south-1.amazonaws.com/chicken

#optinalsudo apt-get update -ysudo apt-get upgrade#requiredcurl -fsSL https://get.docker.com -o get-docker.shsudo sh get-docker.shsudo usermod -aG docker ubuntunewgrp docker

setting>actions>runner>new self hosted runner> choose os> then run command one by one

AWS_ACCESS_KEY_ID=AWS_SECRET_ACCESS_KEY=AWS_REGION = us-east-1AWS_ECR_LOGIN_URI = demo>>  738400679807.dkr.ecr.ap-south-1.amazonaws.com/chickenECR_REPOSITORY_NAME = simple-app

s3cEZKH5yytiVnJ3h+eI3qhhzf9q1vNwEi6+q+WGdd+ACRCZ7JD6

docker build -t chickenapp.azurecr.io/chicken:latest .

docker login chickenapp.azurecr.io

docker push chickenapp.azurecr.io/chicken:latest

1. Build the Docker image of the Source Code
2. Push the Docker image to Container Registry
3. Launch the Web App Server in Azure
4. Pull the Docker image from the container registry to Web App server and run

MLflow

• Its Production Grade
• Trace all of your expriements
• Logging & taging your model

DVC

• Its very lite weight for POC only
• lite weight expriements tracker
• It can perform Orchestration (Creating Pipelines)

This project demonstrates the classification of chicken fecal samples as diseased or healthy using computer vision techniques. The modular structure and the use of pipelines make it easy to follow and reproduce the workflow. The Flask app provides a user-friendly interface for interacting with the classification model.

For more details, refer to the individual implementation files and comments within the code.