0

I'm coding an application that detects vases.
- Vases have several shapes, so I think a way to do it is compare its shape.

  • Firstly: Training, get shape of some vase to test and store in database(I think).
  • Compare: When I put a vase to the camera or capture it, the app will get shape of the vase and compare with existed shape to get the information of the vase.

Does it is the right way?
And I don't know how to get shape of a vase.

Or it have other way?
Please help me with this.
Thanks.

Ngo Van
  • 857
  • 1
  • 21
  • 37
  • hmm.. if you do neural nets, I think you only need to store the weights and not the shapes themselves. – thedjaney May 14 '14 at 05:04
  • The right answer for this is to use an existing Computer Vision library. Because if you need to ask this question, the chance you understand the techniques you'd need to use are almost zero. – Gabe Sechan May 14 '14 at 05:32

1 Answers1

0

I'm not an expert on this matter but, I think this is what you are looking for. How to input the image to the neural network?

Basicall, you translate the image data into something that can be fed into a neural net (link above).

Then, train the network.

Then store the weights of the trained network to be used in the actual app.

Community
  • 1
  • 1
thedjaney
  • 1,126
  • 2
  • 10
  • 28