A Novel Approach for Object Extraction Based On Linear Discriminant Analysis
Sukanya, G., Swarnalatha, S. · Zenodo (CERN European Organization for Nuclear Research) · 2018
Multi view object Extraction plays a major role in tracking and many other applications. Recently different methods are used to extract the object along with boundaries in multi directions. Principal component analysis method is used to extract the object. This method fails due to the high dimensionality and high complexity. So, to overcome the above drawbacks proposed a method called linear discriminant analysis. In this method first extracting the features of the image and converting into the H,S and V planes. k-means segmentation performed to segment the foreground object and extract the boundaries of the object. Experimental results prove to be better and yields better performance when compared to the other state of art methods.