Using a Na"ive Bayes Classifier based on K-Nearest Neighbors with Distance Weighting for Static Hand-Gesture Recognition in a Human-Robot Dialog System
Pujan Ziaie, Thomas J. Muller, Mary Ellen Foster, Alois Knoll · mediaTUM – the media and publications repository of the Technical University Munich (Technical University Munich) · 2008
Abstract. We present an effective and fast method for static hand gesture recognition. This method is based on classifying the different gestures according to geometric-based invariants which are obtained from image data after segmentation; thus, unlike many other recognition methods, this method is not dependent on skin color. Gestures are extracted from each frame of the video, with a static background. The segmentation is done by dynamic extraction of background pixels according to the histogram of each image. Gestures are classified using a weighted K-Nearest Neighbors Algorithm which is combined with a naïve Bayes approach to estimate the probability of each gesture type. When this method was tested in the domain of the JAST human-robot dialog system, it classified more than 93 % of the gestures correctly into one of three classes.