Common Vision Blox - Tools



CVB Manto

Software for recognition and classification of objects that vary

CVB Manto is a versatile software that is able to solve optical recognition tasks not previously possible. The software is based on a specially adapted technology based on support vector machines.

CVB Manto uses all image information from multiple image planes and as such can be applied to monochrome, colour and even multi-spectral images. It automatically identifies the key features that help to identify the object’s class. Once properly trained it can also identify product characteristics that would be very difficult using other techniques, such as different paper types or the ripeness of fruit.


  • Ideal for recognising and classifying organically varying features, i.e. fruits, vegetables, plants
  • Cannot be over trained – the more examples you give, the better the recognition
  • Linear interpolation mode – ideal for dynamic grading with an interpolation of the grade between two fixed grades
  • Special texture mode for identification of textures and organic patterns

Markets and Applications

  • Food and packaging industry
    To create reliable inspection solutions that can deal with the changing organic nature of products
  • Machine builders
    For most difficult applications such as reading stamped or punched numbers on metal or cast surfaces with different degrees of readability and depth
  • Surface inspection
    The analysis of the surface structure (for example recognition of different kinds of wood, classification of herbs and spices or inspection of quality of meat)
  • Traffic technology
    Identification of vehicles by their class, brand or model or for number plate reading
  • Complex OCR tasks like handwriting