In‐situ measurement of dust deposition on bag filters using stereo vision and non‐rigid registration

Abstract

PurposeThe goal is to perform an in‐situ measurement of dust layer thickness on bag filters.Design/methodology/approachPresents a machine vision system in terms of hard‐ and software which is capable of obtaining the relative height difference of a cloth filter surface in different states of operation. The measurement process starts with a stereo reconstruction of filter surface patches. The patches are aligned to form a continuous model using the iterative closest point algorithm. Models acquired at different states of operation are registered using thin plate spline interpolation and afterwards subtracted from each other to obtain the relative height difference.FindingsRelative thickness measurement is possible with an accuracy ranging from 50 to 300 μm, depending on the operational states.Research limitations/implicationsIn‐situ measurement of dust deposition is a new tool for examining the behaviour of commercial bag filters.Practical implicationsProvides the foundation for a system measuring both thickness and density of dust deposition.Originality/valueApplies findings in the field of machine vision to a problem in process engineering and filter design.