Feature Vector Construction of 2D Images using Local and Global Features

Abstract

Object recognition is a process of understanding, design, development and creation of methods to recognize the objects in the image. In this paper the main focus is to create feature vector of 2D images using local and global features of images. Feature extraction is a complex phase in image processing and computer visualization. In proposed method, color image is used as an input image. It transformed into gray-scale image. For feature vector information, local and global features are extracted. Local features are extracted using SIFT method in which key points are identified. Global features are extracted based on the intensity values of images. After that create feature vector using local and global features is high-dimensional. The proposed method is experimented using MATLAB R2012b.