Image Classification

Project Description

Create and train an image classifying model using opencv, SVM and opengridsearch on Jupyter Manual data cleaning for 7 faces

1. Rihanna

2. Amitabh Bachchan 

3. Dwayne Johnson

4. harry Styles

5. Kylie Jenner

6. Leonardo DiCaprio 

7. Steve Harvey

Technologies used in this project, Jupyter notebook, Sklearn for model building, Matplotlib and Seaborn for data visualization, Python, Numpy and OpenCV for data cleaning

Explanation about folder structure files

Dataset: Contains images used to train model

opencv: contains algorithms for face detection

test image: contains images to try opencv and Matplotlin and Seaborn

Celecrity classification.ipynb: contains the code created on Jupyter notebook to create model. Make changes to it as necessary

classdictionary: A specific number is given to each person for model traning

savedmodel.pkl: Trained model-To train your own model with different images you can use the dataset folder and add your images there

Methodology

Steps

1. Data Collection

collect images and separte each person image in separate folders

2. Data Cleaning

use the code to crop the face and create them in cropped folder in dataset. Manually check cropped photos to ensure no inconsistent images. These cropped images will be used to train the model.

3. Training Model

Trained model- using raw images and wavelet transformed images to perform vertical stacking of them and train model and then hyper tune it! Wavelet transformed imaged- to extract features from a cropped image! GridSearch used to try different models with different parameters and come up with best model and parameters

Objective

This project aims is to use machine learning techniques to create a model to identify images that can help businesses. Some of the ways it can be helpful are listed below.

1. Increase Productivity

2. Build Better Content.

3. Prepare for Future Processes.

4. Incorporate in Automation.

5. Assist in Security Threats.

6. Leverage in Hiring Processes.

7. Streamline Customer Processes.

8. Identify Variables in Images.

Note: if following error is given while running code : error: OpenCV(4.6.0) D:\a\opencv-python\opencv-python\opencv\modules\imgproc\src\color.cpp:182: error: (-215:Assertion failed) !_src.empty() in function 'cv::cvtColor'- Delete cropped folder manually if present and rerun code