Perception Based Object Recognition Using SP Theory

Abstract

Improved SP theory is an approach for extracting distinctive invariant features from images. It has been successfully applied to a variety of computer vision problems based on feature matching including object recognition, pose estimation, image retrieval and many others. The main idea is to divide the features extracted from both the test and the model object image into several sub-collections before they are matched. The features are divided into several sub-collections considering the features arising from different octaves that are from different frequency domains. This study provides a preliminary, concise, but complete background of the object detection problem. Problems like more energy consumption, bulkiness, less sophisticated for parallel streams of data and more elapsed time occurred, to resolve this issue combination of IPSO-GA has been used.  Thus, based on this study, for a given problem environment and data availability, a proper framework can be chosen easily and quickly. In this research the accuracy and the capability to detect the object in the noisy medium is achieved.