Abstract
Automatic gesture and pose recognition by machine learning is a significant challenge. This work proposes a simple method for human pose re-cognition using exclusively depth images. The method consists in background removal, feature extraction via histograms of oriented gradients (HOG) and classification through support vector machines (SVM). The method is validated in a test with five human poses, in which a 95.6% success rate is obtained
Users
Please
log in to take part in the discussion (add own reviews or comments).