Image Processing Object Counting

A collection of image processing algorithms including template matching and circle detection.

PythonOpenCVStreamlitTemplate MatchingCircle Detection
Paint Template Matching 1
Paint Template Matching 2
Circle Detection Output
Bird Template Matching 1
Bird Template Matching 2
Poker Template Matching 1
Poker Template Matching 2

Overview

An interactive computer vision web application deployed on Render, built with Python, OpenCV, and Streamlit that automates object identification and counting from digital images.

Features pattern recognition techniques such as template matching and circle detection (Hough Circles / Blob Detection) with customizable threshold controls and color filtering, allowing users to fine-tune detection parameters in real time for precise automated analysis.

Source Code
→
Live Site
→

Features

  • Template Matching: Robust algorithms to find predefined templates within larger scenes.
  • Circle Detection: Utilizing Hough Circle Transform to identify circular objects accurately regardless of noise and variations in lighting.
  • Performance Optimization: Adjust different threshold values to optimize detection accuracy.