Loading presentation...

Present Remotely

Send the link below via email or IM

Copy

Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.

DeleteCancel

Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

HUMAN BODY SKELETON DETECTION & TRACKING

No description
by

Kartik Nagre

on 19 September 2016

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of HUMAN BODY SKELETON DETECTION & TRACKING

HUMAN BODY SKELETON DETECTION & TRACKING

GROUP MEMBERS:
Kartik Nagre
Aditya Gurjar
Anuja Joshi
Rucha Damle
Pune Institute Of Computer Technology,Pune

What we want to achieve...
Our Approach...
Detect And Track
MOTIVATION
To provide utility of devices like Kinect by using a normal camera/ webcam.
To replace use of sensors in hand-held device for motion sensing.
To aim at taking the project working accurately on hand-held devices.
Problem Definition
To create a system using OpenCV to be used as a library to detect human body for all its major joints,generate human skeleton from them and track the generated human skeleton using video stream input through normal camera.
Human Skeleton Stick Model
Detection
Haarcascade - Viola Jones Algorithm
TRACKING
TRACKING
Assumptions for Lucas Kanade :
The pixel intensities of an object do not change between consecutive frames.
Neighbouring pixels have same motion.
Library Architecture
Applications
The main aim of the project is to replace Kinect for most of the tasks it does.
Future Scope
System:
Dynamic learning of all the captured frames(Region of interest) along with Viola Jones and Optical Flow will be more sensible approach.
1. Haar like features are mapped on the object.

2.Conversion into Integral Image for rapid feature value evaluation.
PART - 1
PART - 2
1.AdaBoost Algorithm-select small set of features & train the classifier.
2.Constructs a linear combination of weak classifier to form a strong classifier
PART - 3
1.Construction of cascade classifier in the form of degenerate decision tree.
2.Image/frame having 50 % matching features is passed to next stage
Magic Number (1.618) is used for detection of elbow joint and abdomen joint.
Lucas Kanade Optical flow algorithm
Optical Flow - 2D vector field of displacement vector showing movement of points between consecutive frames
Computation of Lucas Kanade is based on the image brightness constancy assumption.
For motion(u,v) of a point in an image I the brightness of the point does not change,
I(x,y,t)=I(x+u,y+v,t+1)
Pyramidal Lucas-Kanade Algorithm

For large object motions with large dispacement vectors
Gaussian Pyramids built at different levels for two consecutive frames
Motion Gaming
Animation Film Making
Augmented Reality Applications
Dynamic Pose Recognition
Military Applications
Applications:
Deployment of the applications using system on hand held devices.
Sponsored by Rolocule Games Pvt Ltd. Pune
PROJECT...APPLIED
Full transcript