PhD Dissertation:Ibrahim Saygin Topkaya
  • FENS
  • PhD Dissertation:Ibrahim Saygin Topkaya

You are here

VISUAL DETECTION AND TRACKING OF UNKNOWN NUMBER OF OBJECTS WITH NONPARAMETRIC CLUSTERING METHODS

 

 

Ibrahim Saygin Topkaya
EE, PhD Dissertation, 2016

 

Thesis Jury

Assoc. Prof. Dr. Hakan Erdogan (Thesis Advisor)

Assoc. Prof. Dr. Mujdat Cetin

Prof. Dr. Berrin Yanikoglu

Assoc. Prof. Dr. Erchan Aptoula

Assoc. Prof. Dr. Metin Sezgin

 

 

Date &Time: June 2nd, 2016 – 10:00 AM

Place: G025

Keywords : Nonparametric clustering, Dirichlet process mixture models, Chinese restaurant process, multiple object tracking, people counting

 

Abstract

 

Clustering methods that do not expect the number of clusters to be known a priori and infer the number of clusters are known as nonparametric clustering methods in the literature. In this thesis we propose novel approaches to common computer vision applications using nonparametric clustering. We attack the problems of multiple object tracking and people counting. Our main motivation is to approach those as data association tasks where we define the data association problem specific to the nature of the application and benefit from the nonparametric nature of the clustering model. We first propose a detection free tracking method which tracks an unknown number of objects by clustering superpixels. We define the clusters as targets with spatial and visual features and track their changes through time by sequential clustering. The clusters yield tracked targets through time. We also propose a method for clustering short track segments into unknown number of tracks. The clustering similarity is defined using the spatio-temporal features of the short track segments. The clustering process yields robust tracks of objects through time. We use this approach alo to improve the tracking results of the detection free tracking proposed before. Finally we cluster raw person detector outputs to obtain groups of people in a scene and estimate the number of people inside a cluster using the features already extracted for clustering with a proposed metric which is invariant to perspective distortion.