Object Detection and Tracking using Multiple Features Extraction

Bhavya Rudraiah*, ,Dr Geetha K S

doi:10.35940/ijitee.j9437.08101021

Abstract

In most of the video analysis applications, object detection and tracking play vital role. Most of detection and tracking algorithms fail to predict multiple objects with varying orientation. In this paper, the goal is to identify and track multiple objects using different feature extraction methods like Locality Sensitive Histogram, Histogram of Oriented Gradients and Edges. These features are subjected to train classifier that can detect the object of different orientations. Experimental results and performance evaluation depicts the proposed method which uses LSH performs well with an increased accuracy of 98%. This method can precisely track the object and can be utilized to track under different scale and pose variations.

Highlights

Multiple Object Tracking is still an exigent approach as it is challenging due to long term occlusion, multiple targets, slow moving objects, etc
Feature extraction being an important step in object detection and tracking applications, can reduce the complexity and improvise robustness in various environments
In addition to handling specific problems associated with the appearance of the target object, different tracking models represent the target objects in various representation schemes

Summary

Introduction

Multiple Object Tracking is still an exigent approach as it is challenging due to long term occlusion, multiple targets, slow moving objects, etc. Locality sensitive histogram computes histogram by considering all the pixels in an image It takes into account contributions from all the pixels by adding floating point value to each bin. This enables to efficiently track the target object with multiple overlapping regions [3]. Histogram of oriented gradients is a superior feature descriptor which computes the gradient magnitude and angle, performs orientation binning and normalizes the blocks. This reduces the computational complexity and is a well-recognized descriptor for tracking [2]

Methods

Results

Conclusion