Description

ABSTRACT:

            Here we proposed the method to recognize the action in the silhouette of human. Here we extract the BoCP (Bag of correlated posses).  BoCP feature will extract in the sequence of steps. Initially we extract the PCA feature. PCA is stands for Principle Component Analysis. Principal component analysis (PCA) is a mathematical procedure that uses an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values of linearly uncorrelated. After PCA feature extraction we find the correlogram matrix. After extracting the feature we cluster the feature by k-means clustering.   Correlogram is also called as a cross correlation. It is the similarity between observations. It is used to find the repeating pattern. We reduce the correlogram dimension by the use of LDA. We extract training feature for our silhouette dataset by the BoCP feature descriptor. We train this feature by the use of SVM (support vector machine). SVM is the supervised learning models. The training feature will give to the classifier. The classifier will train the feature to predict the result.