Object Recognition Based on Representative Score Features

Abstract

In this paper, we present an approach towards object detection and recognition from various environmental conditions such as foggy morning, dust scenarios, and night vision. The goal of the approach is to develop a holistic feature extraction method over object image patch. To categorize objects, the experimental evaluation has prepared through four classifiers. Investigational results with our own collected video sequences are reported to demonstrate the accuracy of the proposed approach.