Facial Expression Recognition System 🎭

📖 Abstract

This project implements a complete image processing system designed to recognize and classify human facial expressions. By applying an end-to-end computer vision pipeline—from raw image preprocessing to final emotion classification—this system accurately maps facial features to corresponding emotional states (e.g., happiness, sadness, anger, surprise).

🎯 Problem Definition and Objectives

Problem: Accurately identifying human emotions from static images or video frames is challenging due to variations in lighting, facial orientation, and individual physiological differences.

Objectives:

• To build a robust Facial Expression Recognition (FER) system.
• To enhance raw facial images using advanced preprocessing techniques.
• To accurately extract facial features using segmentation techniques.
• To classify the extracted features into distinct emotional categories.

⚙️ Methodology

Our image processing pipeline is divided into three main stages:

1. Preprocessing

Before extracting features, the input images are normalized and enhanced to improve system accuracy. Techniques used include:

• Noise Removal: Applying smoothing filters (e.g., Gaussian or Median blur) to reduce camera sensor noise.
• Histogram Enhancement: Utilizing Histogram Equalization (or CLAHE) to improve the contrast of the facial images, especially in poor lighting conditions.
• Filtration: Using sharpening filters to highlight facial edges and critical regions (eyes, mouth).

2. Segmentation & Feature Extraction

In this phase, we isolate the face from the background and extract meaningful data points:

• Face Detection/Segmentation: Isolating the face region from the background.
• Feature Extraction: Identifying key landmark regions (eyes, eyebrows, mouth) and extracting texture or geometric features (e.g., LBP, HOG, or edge-based features) that correlate with specific expressions.

3. Classification & Clustering

The extracted features are fed into a classification model to determine the final emotion:

• Classification Method: [Insert your chosen classifier here, e.g., SVM, K-Nearest Neighbors (KNN), or a custom clustering algorithm].
• The model outputs the predicted emotional state based on the learned facial patterns.

👥 Team Members

Name	Student ID
[Student 1 Name]	[ID]
[Student 2 Name]	[ID]
[Student 3 Name]	[ID]
[Student 4 Name]	[ID]
(Add more rows if needed. Minimum 4 students per team)

🚀 How to Run

(Instructions on how to setup the environment, install dependencies like OpenCV/Scikit-learn, and run your code will be added here as the project progresses).