0

I need to be able to recognise 3D objects.

The objects - for simplicity consider that the objects are a pen, a coin and a mug.

The environment - the objects are placed on assembly line. The camera is placed above them always at a fixed distance(so the angle should be the same). The light should also be more or less the same.

The goal - the camera spends a few seconds scanning object X and has to be able to say whether it is a pen, coin or a mug.

Ideally the device will be Android tablet but this question is more on the theory and what is the best approach in general. Although the objects will be 3D objects from the real world, they will be completely static during the scan.

Improve this question
2
  • Stackoverflow aim's to help people with programming specific questions. One could write a whole thesis as an answer to your question. You really need to do some research yourself and come back with SPECIFIC questions. Commented Nov 1, 2015 at 12:49
  • I don't need a thesis. I need an advice from someone who has done something similar. Since you are asking...A specific question: Is it a reliable approach to make a number of photos of each object from the same angle and go for 2D image detection? The Vuforia project and a bunch of other frameworks provide such features. Commented Nov 1, 2015 at 13:00

1 Answer 1

Reset to default
1

You can use available tools (e.g. Vuforia) which are capable of 3d object recognition but I'm quite sure you don't need 3d object recognition capability IF the camera is fixed and the real objects would always be recognized from the same angle (more or less). Few images would do the deal, even if the light changes within some limits.

Improve this answer
Sign up to request clarification or add additional context in comments.

1 Comment

Thank you! I will test that option and report with the results.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.