Ask Your Question

The best approach to detect slightly different objects

asked 2018-02-17 05:39:52 -0500

Hey guys,

I just wondering... After saw a thousands of videos about "object detection" I could have an idea of how efficient haar cascada can be be in order to detect objects/face/eye.. .etc

But came to my mind the question: "How can I detect two or more objects that are slightly different of each other?" I can give you an example: WHEELS. Wheels are circular format and basically only the style is different.

Wheel 1:

Wheel 2:

How haar cascade could be efficient to detect those images?


edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2018-02-19 03:34:44 -0500

In that case you have to do a 2 step approach

  1. Start with a general wheel detector
  2. Then make a classifier on the found bounding boxes that classifies the type/style

Good luck!

edit flag offensive delete link more

Question Tools

1 follower


Asked: 2018-02-17 05:39:52 -0500

Seen: 150 times

Last updated: Feb 19 '18