Parts Recognition and Counting System

This repository contains the implementation of a Parts Recognition and Counting System developed for a Pattern recognition course. The system is designed to recognize different industrial parts from images and accurately count the number of parts present using computer vision techniques.

---

Project Overview

Manual stock counting in manufacturing environments is time-consuming and error-prone. This project aims to automate the process by developing a pattern recognition system that:

• Identifies different types of parts from images
• Assigns labels to recognized parts
• Counts the number of parts present in an image

The system is implemented and demonstrated using a Jupyter Notebook.

---

System Workflow

1. Image input (captured using mobile phone)
2. Image preprocessing
3. Feature extraction
4. Part recognition using a machine learning model
5. Parts counting using image processing techniques
6. Output of part label and quantity

---

Key Features

• Supports recognition of multiple part types
• Handles images with varying lighting, angles, and backgrounds
• Counts parts arranged in different configurations (scattered, grouped, overlapping)
• Uses open-source computer vision libraries for processing and counting

---

Techniques Used

• Image preprocessing (normalization, grayscale conversion, noise reduction)
• Feature extraction techniques (e.g., PCA or similar methods)
• Machine learning–based classification
• OpenCV-based parts counting algorithms

---

Dataset

• Multiple images per part with different:
  • View angles
  • Lighting conditions
  • Backgrounds
• Additional images containing multiple parts for counting

---

How to Run

Step 1: Create and Activate a Virtual Environment (optional)

Open Command Prompt in the project directory:

python -m venv venv
venv\Scripts\activate

---

Step 2: Install Required Dependencies

pip install --upgrade pip
pip install -r requirements.txt

---

Step 3: Run the Jupyter Notebook

jupyter notebook

Then open and run:

Machine Part recognition and counting (training and tkinter).ipynb

Run the cells sequentially to reproduce the results.

---

Learning Outcomes

• Practical application of pattern recognition concepts
• Experience with image-based feature extraction
• Implementation of object recognition and counting algorithms
• Handling real-world image variations and noise

---

Notes

This project is developed for academic and learning purposes only. The implementation focuses on demonstrating pattern recognition techniques rather than production-level deployment.