Available Student and Research Projects

(Difference between revisions)
Jump to: navigation, search
m (→Micro-Manipulation)
m
Line 35: Line 35:
 
===External Links===
 
===External Links===
 
* [http://www.hadleyweb.pwp.blueyonder.co.uk/ CombineZ]
 
* [http://www.hadleyweb.pwp.blueyonder.co.uk/ CombineZ]
−
 
−
==Automated photo stitching==
 
−
{|align="center"
 
−
|+ '''Hilbre Island input images'''
 
−
|-
 
−
|[[Image:hilbreisland1.jpg|100px]]||[[Image:hilbreisland2.jpg|100px]]||[[Image:hilbreisland3.jpg|100px]]||[[Image:hilbreisland4.jpg|100px]]||[[Image:hilbreisland5.jpg|100px]]||[[Image:hilbreisland6.jpg|100px]]||[[Image:hilbreisland7.jpg|100px]]
 
−
|-
 
−
|}
 
−
 
−
{|align="center"
 
−
|+ '''Resulting panorama image created with [http://hugin.sourceforge.net/ Hugin]'''
 
−
|-
 
−
|[[Image:hilbreislandpanorama.jpg|520px]]
 
−
|-
 
−
|}
 
−
 
−
===Premise===
 
−
* A set of images taken with the same camera-settings (aperture, exposure time, focal length) and center of projection but different viewing directions (mainly yaw, pitch)
 
−
* Manually selected correspondences.
 
−
 
−
===To Do===
 
−
<math>
 
−
\lambda\,\begin{pmatrix}m^\prime_{1}\\m^\prime_{2}\\f\end{pmatrix}=
 
−
\begin{pmatrix}h_{11}&h_{12}&h_{13}\\h_{21}&h_{22}&h_{23}\\h_{31}&h_{32}&h_{33}\end{pmatrix}\,
 
−
\begin{pmatrix}m_{1}\\m_{2}\\f\end{pmatrix}
 
−
</math>
 
−
 
−
* Improve the correspondences using 2D cross-correlation (roll-angle will assumed to be low).
 
−
* (Re)implement the [http://panotools.sf.net/ panotools]-solver preferably in [http://www.ruby-lang.org/ Ruby]. Use given correspondences to optimize camera-parameters: rotations, common focal length.
 
−
* Adopt existing method for removing fringes (transitions/blending)
 
−
Optional:
 
−
* Automatically find correspondences
 
−
* Optimize more camera-parameters (distortion,...)
 
−
 
−
Also see [[Panorama Viewer|panorama viewer]].
 
−
 
−
===External Links===
 
−
* [http://www.ptgui.com/ Photo stitching software PTGUI]
 
−
* [http://en.wikipedia.org/wiki/Rotation_matrix Rodriguez Matrix]
 
−
* [http://en.wikipedia.org/wiki/Panotools Panorama Tools software suite] ([http://www.path.unimelb.edu.au/~dersch/ mirror of Helmut Dersch's former web-page])
 
−
* [http://hugin.sourceforge.net/ Hugin]
 
−
* [http://www.janrik.net/ptools/ExtendedFocusPano12/index.html Depth of focus for panorama tools]
 
−
* [http://www.cs.ubc.ca/~mbrown/autostitch/autostitch.html Autostitch (commercial)]
 
  
 
==RANSAC==
 
==RANSAC==
Line 90: Line 47:
 
* Extend RANSAC implementation to more degrees-of-freedom problem (maybe use line- and point-features)
 
* Extend RANSAC implementation to more degrees-of-freedom problem (maybe use line- and point-features)
  
−
==HDR imaging==
 
−
{|align="center"
 
−
|+ '''Stanage Edge'''
 
−
|-
 
−
|[[Image:stanage1.png|160px]]||[[Image:stanage2.png|160px]]||[[Image:stanage3.png|160px]]
 
−
|-
 
−
|}
 
−
{|align="center"
 
−
|+ '''Processed image after composing HDR and tonemapping'''
 
−
|[[Image:stanage.png|240px]]
 
−
|-
 
−
|}
 
−
 
−
Merging pre-aligned 8-bit colour photos into a [http://en.wikipedia.org/wiki/High_dynamic_range_imaging High dynamic range] image
 
−
will be of high importance as long as low-cost HDR-cameras are not available for the consumer market.
 
−
===Premise===
 
−
* A pre-aligned exposure-series of pictures
 
−
===To Do===
 
−
* Detect over- and undersaturated colour-components
 
−
* Estimate camera sensivity
 
−
* Merge images into a single HDR image
 
−
Optional:
 
−
* Map to reasonably realistic-looking low-range image
 
  
 
===See Also===
 
===See Also===
Line 120: Line 54:
 
* [http://qtpfsgui.sourceforge.net/ HDR workflow with Qtpfsgui] to create HDR images and tonemapping
 
* [http://qtpfsgui.sourceforge.net/ HDR workflow with Qtpfsgui] to create HDR images and tonemapping
 
* [http://wiki.panotools.org/HDR_workflow_with_hugin HDR workflow with hugin] to create HDR panoramas
 
* [http://wiki.panotools.org/HDR_workflow_with_hugin HDR workflow with hugin] to create HDR panoramas
−
 
−
 
  
 
==Micro-Manipulation==
 
==Micro-Manipulation==

Revision as of 17:54, 8 June 2007

Steel gates of entrance to City Campus next to Sheaf Building

Contents

Student Projects

We always offer projects to motivated students (first degree, Masters of Science, ERASMUS student, ...). If you are interested in doing a project in computer vision, let us know. Here is a description of ongoing research areas to give you an idea of possible projects. You can also suggest a research topic yourself.

If you work with us, you can learn a lot of skills which are relevant for a career as a software developer:

  • Computer Vision, Signal processing, Robotics
  • Linear Algebra, Analysis
  • Software Engineering

We are using state-of-the-art cross-platform software tools:

  • Source-code documentation with Doxygen logo.png doxygen
  • Cross-platform user-interfaces with Qt logo.png Qt. You can develop full-featured GUI-software which runs under Tux.jpg GNU/Linux, Ms-windows logo.png Microsoft Windows, and Macos.gif MacOS!
  • Platform-independent Stl logo.gif Standard Template Library
  • Platform-independent C--boost logo.gif Boost Library
  • Scripting using the dynamically-typed object-oriented programming language Ruby.png Ruby

Project areas

Stitching for microscopes

A bird's feather (reflected light, darkfield) (7.2 MByte video, 10.1 MByte video)
Stitching using the feedback of the microscope's drive

Premise

  1. A microscope-video of an object being moved in x-, and y-direction (parallel to the focussed plane)
  2. Later a microscope-video of an object being moved in x-, y-, and z-direction (i.e. including depth changes)

To Do

  • Generate stitched image from the input-video (linear complexity desirable) without feedback from microscope-drive
  • Cross-compare images to avoid a drift of the estimated shift
  • Later provide extended depth of field by maximising a focus measure.

See Also

External Links

RANSAC

Recognition and tracking with three or four degrees-of-freedom. More ...

Random sample consensus is a method for object recognition. This project is about recognising macroscopic rigid objects (e.g. household and office articles like cups, stapler, ...

To Do

  • Select point-features and a suitable similarity measure
  • Implement RANSAC algorithm and apply to at least 3 degrees-of-freedom problem.
  • Demonstrate algorithm on real object.

Optional

  • Extend RANSAC implementation to more degrees-of-freedom problem (maybe use line- and point-features)


See Also

Micro-Manipulation

Micro manipulator picking up a crumb of ground coffee

This project is about manipulating objects which can be seen under a microscope. The size of the objects typically is up to about 750 micrometers.

Premise

An optical microscope with a motorized stage and a low-cost firewire video camera. There is an early prototype of a gripper mounted on a microtranslation stage. Parts with limited accuracy can be manufactured using rapid prototyping or in the lab.

To Do

The task is to construct and build a more advanced gripper. Possible ideas are:

  • Use strain gauges to have feedback
  • Develop gripper with more degrees of freedom

See Also

External Links

See Also

See Also

External Links

Retrieved from "https://vision.eng.shu.ac.uk/mmvl/mmvlwiki/index.php?title=Available_Student_and_Research_Projects&oldid=1767"
Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox