-1

I have pixel by pixel based labeled (5 label classes 0, 1, 2, 3, 4) data of around 200,000 images. Each image sized 240x240. Can anyone suggest me how can I train this data efficiently and what will be the best possible classification method for it?

Thank all in advance for help.

  • What have you tried so far? Where are you getting confused? There are lots of tutorials out there - have you searched at all? – Tchotchke Nov 07 '16 at 17:58
  • I searched lot of methods and before that I have done classification of each image to classify them in different categories. But here the case is different as each pixel (ground truth) is labeled with specific class. I want to have optimized solution for Training based on each pixel class and than test it to check the performance. Can you suggest something? – Ghazanfar Latif Nov 07 '16 at 22:22

1 Answers1

0

You can use Convolutional Neural Network. You can see this article http://www.pyimagesearch.com/2016/08/01/lenet-convolutional-neural-network-in-python/

Rabindra Nath Nandi
  • 1,433
  • 1
  • 15
  • 28