The measurement of dirt count using a scanner-based image analysis system

The existing TAPPI standard methods for measuring dirt in pulp (T213) and i n paper and paper board (T437) have proven to be tedious, impractical and u nreliable due to their subjectivity. The need for more stringent quality control of pulp cleanliness in the fine paper sector, coupled with the ever-increasing use of recycled fibre, has forced the Australian paper industry to look for a more reliable and faster method for measuring dirt. The use of image analysis for off-line testing has often been suggested to improve objectivity and to substantially increa se the speed of measurement. Against this background, an in-house software system utilising the latest scanner-based image analysis technology was dev eloped at Amcor Research and Technology (ART), in:conjunction with Royal Me lbourne Institute of Technology to measure dirt count in samples of pulp an d paper. A modified version of the provisional TAPPI dirt count method (T56 3) was used as the basis of the algorithm developed for this application. The method has proven to be quick, reliable:and repeatable for measuring di rt in laboratory, research and mill applications. This paper outlines diffi culties with existing systems, the refinement of these methods and some cas e studies based on typical applications for dirt count.