Wood cellular structure evaluation using image analysis methods

Light microscopy images of wood cellular structure are used to study physical and biological characteristics of the wood. Information of the microscopic structure can be used in dendrological studies, evaluating density and porosity and even determining physical changes of the modification process. Microscopic images can also be used to identify different wood species. In many cases, the microscopic images are analyzed manually which is very time consuming (10-30 min per image) and usually it is not possible to include all the cells from the image into the analysis. Therefore, this study shows an implementation of fast image analysis based wood cell structure analysis method which can be used for screening large sets of microscopy images automatically. Accuracy of the method was evaluated against manual measurements and the average difference varied from 1.1 % to 6.4 % for cell diameter and 5.2 % for cell wall thickness determination depending of the used parameters. Processing time for the images varied from 3 to 60 seconds depending of the image resolution and the number of analyzed cells.