Utilizing machine vision in wood chip quality analysis

Jyrki Raitila, Juho Riekkinen

    Research output: Contribution to conferenceConference articleScientific


    Machine vision technology for sorting of fuel wood by quality and particle size, as well as for assessing the volume of a delivered fuel wood load was tested by VTT and JAMK. INFRES partners provided chip samples from all over Europe for testing. RGB images proved to work very well when identifying shapes and sizes of chips. If odd particles have the same colour as woody material, RGB images could not identify them. Measuring wood chip loads with a time-of-flight (TOF) camera rendered the most promising results. The average error was about 10%. Compared to visible light technology, near infrared (NIR) spectroscopy proved to be much more accurate in determining fuel wood moisture content and detecting foreign objects. Technology based on visible light is not able to work online (moving chips). To the contrary, NIR technology proved to work online and therefore could be used at a power plant or fuel wood terminal where wood chips are moved with a conveyer. However, NIR technology has other challenges such as not being able to give reliable moisture information with regard to frozen materials. Furthermore, an outlook on future research needs was provided.
    Original languageEnglish
    Publication statusPublished - 2013
    MoE publication typeNot Eligible
    Event46th International Symposium on Forestry Mechanization, FORMEC 2013 - Stralsund, Germany
    Duration: 28 Sept 20131 Oct 2013
    Conference number: 46


    Conference46th International Symposium on Forestry Mechanization, FORMEC 2013
    Abbreviated titleFORMEC 2013


    • wood chips
    • machine vision
    • moisture content


    Dive into the research topics of 'Utilizing machine vision in wood chip quality analysis'. Together they form a unique fingerprint.

    Cite this