PubAg

Main content area

Towards the usage of image analysis technique to measure particles size and composition in wood-polymer composites

Author:
Michalska-Pożoga, Iwona, Tomkowski, Robert, Rydzkowski, Tomasz, Thakur, Vijay Kumar
Source:
Industrial crops and products 2016 v.92 pp. 149-156
ISSN:
0926-6690
Subject:
algorithms, compliance, computer analysis, computers, extrusion, functional properties, image analysis, injection molding, polymers, wood, wood flour
Abstract:
Polymer composites are composed of different materials that have been found to affect the different functional properties of the obtained products. One of the most interesting example in this field is wood-polymer composites (WPC), where the solid phase is composed of different origins and wood flour. Regardless of the dimensions of wood particles, they have been found to play a significant role on determining the overall properties of the resulting WPC material. Therefore, determination of wood particles size as well as the ability to determine the changes in granulometric composition induced during processing (injection molding or extrusion) are of the special significance. Methodology of determination, even with the use of computer systems for image analysis, is rather complicated and requires a lot of time, and despite that the results obtained are not always reliable. Therefore, it is reasonable to seek for better solutions allowing to obtain fast and reliable results. So in this article, we present a highly viable method of wood particles size multi-stage determination using image analysis and numerical computation program. The procedure of grain size image analysis includes a step of preparing sample images to the computer analysis, it enters images binarization, morphological operations, determination of objects overlapping and determination of its dimensions. The developed numerical algorithm for the image analysis allows to set the parameters of compliance to match the shape extracted from the image composite particles to their real shapes.
Agid:
5496573