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

Bin Picking system with Kinect Camera

This project explores the use of a Microsoft XBOX Kinect Camera instead of the conventional LASER Camera
by

Chintan Mishra

on 15 November 2012

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Bin Picking system with Kinect Camera

Chintan Mishra
Zeeshan Ahmed Khan Bin Picking With Kinect Introduction
KINECT
What Why Where
Structure & Specification
PCL
3D Perception
Challenges
Future Work INTRODUCTION Bin-Picking
Laser Camera
Major Industrial Activity
Profitable Approach
Safe
Efficient & Cost Reduction
Depth Camera KINECT What Why Where
Structure
Specifications
Limitations What
Gaming Console By Microsoft
Released Nov 2010
Project NATAL
Why
To Play Games without touching game controller
Where
Microsoft XBOX WHAT WHY WHERE PCL STRUCTURE 3D DEPTH SENSORS RGB CAMERA MOTORIZED TILT MULTI-ARRAY MIC Kinect Structure SPECIFICATIONS Field of View (Horizontal, Vertical, Diagonal)58° H, 45° V, 70° D
Depth Image 640x480 @ 11-Bit (100Mbps), 30Hz
Colour Image VGA 640x480 @ 8-Bit
Range ~0.8 m to ~4 m Point Cloud Library
Open Source Software for object recognition algorithms
Collection of modular C++ libraries (keypoints, features, surface etc.)
Kinect Support (Plug & Play) Calibration
Data Collection
Matching
POSE Estimation 3D PERCEPTION
CALIBRATION Not mandatory as the reference frames are almost the same for the depth and RGB cameras

Mandatory to adjust the lighting conditions
DATA COLLECTION Data Acquisition
XYZ, Colour
Filtration
Segmentation
Outlier Removal
Storage
*.pcd format POSE ESTIMATION Align the reference model to the instance in the scene (position estimation)
Determine fitness score
Estimate position of the object
Transformation to robot coordinate system
MATCHING Extract keypoints from reference model and scene
SIFT, Uniform Sampling
Compute descriptors
SHOT, Spin Images, PFH
Find valid correspondences
Euclidian distance CHALLENGES Descriptiveness of the training model
Cluttered scene
Lighting
Reflective surfaces
Importing to Meshlab and scaling the model
Kinect Scanning
Surface reconstruction Q & A FUTURE WORK Estimate exact position of object
Calculate the gripping points
Optimal path planning for robot arm
Collision avoidance during grasping LIMITATIONS Can see through Glass (because of Infrared)
Cannot work in Sunlight/Outdoor (because of Infrared)
Does not capture detailed description of small objects.
Full transcript