English | Türkçe

Object Detection with OpenCV

Image Processing and Computer Vision Documentation Project (EN, TR)

Examples

There are three examples in the repository.

  1. Haar Cascade - Object detection face and eye etc.
  2. Color Detection - Object detection and tracking using object color.
  3. Template Matching - Object detection with template matching.
  4. Deep Learning - Object detection with deep neural network (DNN).

Example 1: Face And Eye Detection

Source code location: src/FaceAndEyeDetection/

Object detection examples with haar cascade classifier algorithm (Face, eyes, mouth, other objects etc.). Cascade Classifier Training http://docs.opencv.org/3.1.0/dc/d88/tutorial_traincascade.html

What is Haar cascade? Haar cascade classifier Object Detection using Haar feature-based cascade classifiers is an effective object detection method proposed by Paul Viola and Michael Jones in their paper, "Rapid Object Detection using a Boosted Cascade of Simple Features" in 2001. It is a machine learning based approach where a cascade function is trained from a lot of positive and negative images. It is then used to detect objects in other images.

Requirements

Face and eye detection by the camera using haar cascade algorithm.

Video:

Example 2: Object Detection and Tracking Using Color

Source code location: src/ColorBasedObjectTracker/

An example of an application where OpenCV is used to detect objects based on color differences.

Requirements

Example 3: Object Detection with Template Matching

Source code location: src/TemplateMatchingObjectDetection/

Template matching is a technique for finding areas of an image that match (are similar) to a template image (patch).

Requirements

My blog post for template matching.

Example 4: Object Detection with DNN

Source code location: src/DeepNeuralNetwork/

In this tutorial you will learn how to use opencv dnn module for image classification by using MobileNetSSD_deploy trained network. My blog post for deep neural network.