#
Page tree
Skip to end of metadata
Go to start of metadata

Motivated from spectroscopic work in laboratories, an approach to unscramble spectra into their principle components is employed.

 

Function

An unscrambling process is controlled by the maximum variance information in selected spectra. The perception gained is composed of score values of calculated principle „patterns“ in the spectra. By assigning components to colour channels (which hold application relevant information), a perception is obtained based on principle components.

User Interface

Principle components are extracted unsupervised from selected spectra.
The value per component is shown through a score image at the bottom right of the Model perspective.
By comparison of the score images to expectations and assignment to color channels, a CCI image is obtained (shown to the right of the data view)

Work Flow

  • Select spectra from the scene which potentially holds information of interest.
  • Inspect gained component score images (match with your expectations).
  • E.g. apply pre-processing and in parallel compare results with your expectations
  • E.g. crop the scene to focus only on objects of interest.
  • E.g. set the scaling to selected spectra to arrange the colour contrast in the result
  • Assign those components to colour channels which best describe application relevant information.
  • Save the model for later usage (e.g. to setup the live streaming)

Hints

Use the Extract method interchangeably with hyperspectral preprocessing. Try to find a preprocessing situation where spectral information of interest is shown as distinct variance information.
Often this is found by switching through the preprocessing methods and the comparison of results with expectations.

Pro/Con's

(plus) the extraction process doesn't necessitate knowledge about observed objects.

(plus) the component extraction is done in a statistically well validated way.

(plus) suggested as standard analyzing step - to get experience about the spectral composition

(minus) might fail when components of interest are interfered by much stronger side-components such as noise.

Typical application

  • as analysing tool
  • suggested as standard analysing step

Mentioned in:

Found 4 search result(s) for "extract method".

Page: Impurity detection on sausage (Manuals)
... application relevant information. If no application relevant information is available from the score images, the Extract method might be not suitable for your application. In this case, continue with one of the other CCI methods ... ...
Oct 10, 2019
Page: Select Spectra (Manuals)
... Numerous functions in the Perception Studio program necessitate spectra to be the input. Like the CCI Extract method which unscrambles selected spectra into principle patterns. The Select Spectra Tool enables ... ...
Oct 10, 2019
Page: Model Hyperspectral Information (Manuals)
... Preview Method: unsupervised method showing spectroscopic differences in different colors. CCI Extract Method: extraction based on variance information. CCI Correlate Method: correlation of pure substances CCI Constrain Method: constraining colors CCI Design Method: design ...
Oct 10, 2019
Page: Hyperspectral imaging step by step (Manuals)
... Model perspective → Select the data for the modelling process. → Choose from one modelling method available in the ribbon. → Develop a model and save it for later usage. Dependent on the application different ...
Dec 11, 2019



© 2019 by Perception Park GmbH
The content of this page and any attached files are confidential and intended solely for the addressee(s). Any publication, transmission or other use of the information by a person or entity other than the intended addressee is prohibited. If you receive this in error please contact Perception Park and delete copied material. Perception Park GmbH, Wartingergasse 42, A-8010 Graz; Austria; FN 400381x

  • No labels