Ask Your Question
1

Do I need to write a logic of pattern recognition (classification of objects) or OpenCV already have such a function?

asked 2012-10-05 18:13:41 -0600

Anderson gravatar image

I know that the OpenCV has the tools to do noise filtering, edge detect. But as i know it only preparation for recognition . Pattern recognition mean classification objects (the selection one object from others). Should I write your own algoritm of recognition (neural network for example) or OpenCV already have such a function? OpenCV works with face recognition only or can be set to any object recognition (after learning for example)? What algorithms are used for this?

edit retag flag offensive close merge delete

2 answers

Sort by ยป oldest newest most voted
5

answered 2012-10-05 20:18:26 -0600

OpenCV is a wonderful thing my friend :)

Take a look at the Machine Learning section of the library, it contains implementations of many state-of-the-art machine learning and pattern recognition techniques, you don't need to reinvent the wheel. And the interface is pretty simple: MachineLearningMethod::train() and MachineLearningMethod::predict()

You can find implementations for:

  • Statistical models
  • Normal Bayes Classifier
  • K-nearest neighbors
  • Support Vector Machine
  • Decision trees
  • Random forest
  • EM
  • Neural Networks
  • ...

It doesn't work only with face recognition, if you provide the training data you can learn anything.

Happy pattern recognition!

edit flag offensive delete link more
4

answered 2012-10-09 01:29:37 -0600

Kirill Kornyakov gravatar image

Here are a couple of tutorials to start:

  1. Face and eyes detection.
  2. Detection of a known planar object.
  3. Template matching.
  4. Detecting lines and circles.

So, you can see that there are a number of methods for particular tasks. And Martin is right, you should also study the opencv_ml module.

edit flag offensive delete link more

Question Tools

1 follower

Stats

Asked: 2012-10-05 18:13:41 -0600

Seen: 1,257 times

Last updated: Oct 09 '12