Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 13

CENTRAL UNIVERSITY OF 

HARYANA 
(Department of CS & IT)

Master of Computer Applications


Presentation on :
Particle Filter
(Computer Vision)
Submitted To Submitted BY
Dr. Suraj Arya   Hemlata
( Assistant Professor)  MCA(4th sem.)
(Department of CS & IT) Roll no- 200581
Content:-

 Introduction
 What is Particle Filter?
 Application of Particle Filter
 Recursive Bayesian Estimation
 Monte Carlo method
 Why Particle Filter ?
 Pros & Cons of Particle Filter
Introduction:-

Particle Filters : two words

Filter : a procedure that estimates parameters (state) of a system.


Particles : a set of randomly chosen weighted samples used to
approximate a pdf(Probability Density Function) .

Estimation:-
Estimation is the process by which we infer the value of a quantity of
Interest.
Application of State estimation /filtering

State estimation or filtering has many


applications

1) Predicting economical data


2) Tracking of aircraft positions from radar
3) Tracking of people or cars in surveillance videos
4) Mobile robotics etc...
What Is A Particle Filter ?
Particle filter is :
1. A technique for implementing recursive Bayesian filter by Monte Carlo
sampling.
2. Particle filters methods are recursive Bayesian filters which provide a
convenient and attractive approach to approximate the posterior
distributions when the model is nonlinear .

Posterior density:-the distribution of a set of unknown parameters conditional on


the current data.
Monte Carlo Methods:-
 Monte Carlo methods are a broad class of computational algorithms that rely on repeated 
random sampling to obtain numerical results.

 They are often used in physical and mathematical problems and are most useful when it is
difficult or impossible to use other approaches.

 Monte Carlo methods are mainly used in three problem classes


1). Optimization
2) Numerical Integration
3) Generating draws from a Probability Distribution..
Recursive Bayesian Estimation

1). Recursive Bayesian estimation, also known as a Bayes filter.

2). Recursive Bayesian estimation is an approach used in statistics and machine


learning (ML) that estimates the current state of a system. The framework is used
in robotics and automotive technology, where machines are taught to perform a
task that requires estimation.
Example:-
A self-driving car can estimate its location using the recursive Bayesian
estimation framework. The car obtains its starting position via a Global
Positioning System (GPS). Its built-in algorithms that help it estimate its current
location after a certain amount of time or distance. The algorithms often use the
Bayesian mathematical concept in statistics.
Applications Of Particle Filter
Medical Image Analysis:-
Particle filters have been widely used in medical image
analysis
⇒ contour tracking, image segmentation
Continue.....

Human Computer Interaction:-


Particle filters are used in HCI for tracking
⇒ face and hand tracking can be coupled with facial expression and
gesture recognition.
Continue....
Image Restoration:-

Particle filters have been used for image enhancement and denoising
Key Idea:-

They are commonly based on the idea that the data distribution is unknown
and that distribution “particles” or samples are evaluated and then analyzed
and aggregated into more meaningful results to find the minimum of the cost
function
Pros & Cons of Particle Filter:-

Pros :-

1) Estimation of full pdfs

2) Non linear dynamic systems

3) Can be parallelized

Cons :-
1) Degeneracy problem

2) High number of particles needed

3) Can be computationally expensive


Thank You

You might also like