Object Recognition using Microsoft Kinect SDK 1.7

Question

I am trying to recognize different objects using Microsoft kinect 360. I am using Microsoft kinect sdk 1.7,OpenCV 2.4.6 and visual C++ for this purpose. Most of the object recognition techniques are based on training data. But I want to build the system so that it can be used for any generic purpose not only for objects it has been trained. The system should have some intelligence to do that task. For doing so, is there any algorithm or machine learning technique available?

Toby Sharp Toby Sharp · Accepted Answer · 2013-11-18T10:39:19

If you want the system to be able to recognize any object at all, you would need to train it on the whole world, and if it worked you would have invented something groundbreaking. Object recognition scenarios work by having a training database, where the images in that database are used to inform the system about how to discriminate between the different object classes.

Object Recognition using Microsoft Kinect SDK 1.7

1 Answers