Biologically Motivated Feature Extraction And Object Categorization

Pattern Recognition is also called Pattern Classification representing a process and analysis to a variety of numerical,literal and logical information. The aim is to describe and explain several phenomenon. Pattern Recognition is also an important part of information science and artificial intelligence . Exploring the processing mechanism of human implementing Pattern Recognition task is to solve vision task. At present, the above missions are also hot problem in the field of pattern recognition and computer vision. So, the emphasis of this paper is to research the meaning and practical value of biologically motivated visual mechanism feature extraction and object categorization.This paper will analyze the main principle of biologically motivated visual cortex. Additionally, the design and implementation of feature extraction and object classification. The major contents of this paper are:(1) The first part of the paper detailedly introduces hierarchy architecture of primate visual cortex including the processing course of image information. Beyond this, architecture of algorithm,the processing of biology physiological reaction and implementation foundation are all described as well.(2) Facing object categorization task, we significantly research feature extraction and classifier designing motivated by biology visual cortex. At the aspect of feature extraction, we utilize image processing algorithm to imitate processing function of visual cortex including relevant research on the current situation and development of Gabor filter. Based on two-dimension Gabor function having the optimization of time domain and frequency domain and owning perfect selectivity of direction and frequency, this paper implement a kind of pattern recognition based on two-dimension Gabor filter. at the aspect of classifier designing, due to SVM having the advantage of Global optimization,simple structure,popularization, it is especially applicable to small sample and multi-dimension feature. So, we choose SVM as classifier to categorize and recognize.(3)Describing the software development platform and implementation of the Imitated biology visual cortex algorithm model. Finally this paper Introduces some basic data structures and designing philosophy. The core job is elaborating the structure and frame of algorithm including a variety of functional module of implementation and algorithm philosophy.