Examples

Classification by One Feature

  • Size:

Classify objects based on their size.

size1 size2 size3

  • Color:

Classify objects based on their color.

color1 color2 color3

  • Contour:

Classify objects based on their contour.

contour1 contour2 contour3

  • Patterning:

Classify objects based on their patterning.

patterning1 patterning2 patterning3


Classification by Several Features

Classify objects based on several features. For this example size, color and patterning were used.

multifeature1 multifeature2 multifeature3


Complex Background

Our software can also deal with complex backgrounds.

In this example all sunflower seeds are treated as background and thus are not transformed to objects. Only the pieces of wood are treated as objects.

background1 background2 background3


Combination of several Sensors

It is also possible to combine the information obtained by several sensors in order to enhance the final results. Any combination of sensors can be used (e.g. 2 RGB cameras + 1 hyperspectral camera + 1 metal detector).

In this example, the interim result of the RGB camera (detection of the wooden plate on the belt conveyor) is combined with the interim result of the hyperspectral camera (detection of loose knots and resin pockets). Thus, the exact location of the defects on the wooden plate can be found.

RGB Camera

RGB Camera

Hyperspectral Camera

Hyperspectral Camera

sensorcombination3 sensorcombination4