An Optimized Method for Object Detection and Tracking Using Hybrid Filtering

Abstract

The field of object detection gain a wide popularity due to its various number of applications. Object detection and tracking are majorly used in traffic detection and monitoring and for medical and various other fields. In this work an object detection and tracking algorithm is proposed on basis on blob analysis and hybrid filtering for tracking. The proposed work is based on usage of two major filters, particle and Kalman filter. The work is done on available data set, where the samples of various objects are considered like vehicles, pets and animals. The proposed method will work on every moving object so it can be applied on any other object too. The work done had shown a significant improvement in existing literature. The proposed algorithm almost gains an accuracy of more than 92% as average for various samples. In future some machine learning technique can also be included to enhance further detection.