Networking xylarium database toward novel wood anatomy by computer vision
|Ratih Damayanti (Forest Product Research and Development Center, Indonesian Ministry of Environment and Forestry)
|Esa Prakasa (Research Center for Informatics, Indonesian Institute of Sciences)
Danang Sudarwoko Adi (RISH, Kyoto University)
Hairi Cipta (RISH, Kyoto University)
Application of promising novel technique that is under development in RISH would certainly open new domain to explore wood diversity. Tropical forest is a wealth of biological resources, thus developing networking on xylarium database by computer vision between RISH and Indonesia will support the objectives of RISH in global environment protection research, harmonized utilization of wood resources and enhancing international collaborative research towards these objectives. A project ‘Networking xylarium database toward novel wood anatomy by computer vision’ was being run since 2017. In 2–9 September 2017, short visit to Kyoto University was carried out by researchers from Indonesia. Some laboratory activities collaborated with students from Kyoto University have been conducted including taking macroscopic and microscopic images of tropical wood species and 15 species of bamboo for identification using computer vision. Discussion in computer vision techniques developed by RISH and Indonesia also has been carried out. The project then continues to add with another 16 species of Indonesian bamboo (totally 31 species), 47 species of Shorea-Dipterocarpaceae, 13 species of tropical timber, amd 18 genera of Fagaceae. The results are in process of data processing for publication. This report focuses on computer vision technique developed using combination of wavelet and LBP (Local Binary Pattern) on three tropical timber species: Neolitsea sp., Horsfieldia glabra and Sterculia cordata.