Related items
Showing items related by metadata.
-
DocumentGrowth estimation of standing trees using artificial intelligence
XV World Forestry Congress, 2-6 May 2022
2022Also available in:
No results found.Growth assessment in an ecosystem is an essential element in management and decision making. Such an exercise helps in development and biodiversity management in a natural ecosystem. The assessment process, however, demands time and manpower. Developing an automated tool not only helps in saving the above-mentioned resources but also in expanding the area of coverage for assessment. We are developing an artificial intelligence based tool using image data for growth assessment. The method will be demonstrated in plantations of eucalyptus and teak. The plantations are established in straight lines using a single species of trees. It can be assumed that the trunk texture of these trees is similar while the shape can be different. Estimating the standing timber volume is important to assess the growth, harvestable timber volume, and plan on the transportation logistics of harvested timber. Every tree must be manually measured in the existing method of volume calculation demanding time and manpower. These costs can be cut down while maintaining the accuracy using images processed with statistical learning methods such as Convolutional Neural Network. The plantations will be partitioned into grids and digital images will be taken from the edges of this grid. These RGB digital images will be processed to determine the growth parameters such as girth at breast height, height, and tapering of the trees. Transfer Learning is to be used in modifying the existing neural network in identifying 3D shapes of individual objects from 2D images, multi-spatial depth estimation, and volume determination. A cost-effective automated tool to estimate the timber volume of standing trees in real-time will be developed. While estimating the volume by this method, a significant amount of time and manpower can be saved without compromising the accuracy compared to the conventional method. Keywords: Monitoring and data collection, Adaptive and integrated management, Innovation, Policies, Research ID: 3621691 -
ArticleSMART Global Ecosystems an academic-industrial partnership to integrate artificial intelligence in forestry
XV World Forestry Congress, 2-6 May 2022
2022Also available in:
No results found.SMART Global Ecosystems is an alliance between the University of Valladolid and the technological company SNGULAR to promote research and training in the area of artificial intelligence and sustainable environmental data science through information and communication technologies (ICTs) and big data analysis. ICTs, together with massive data analysis, open a world of opportunities to implement training, research, and social awareness actions to develop data science and artificial intelligence methods and products that facilitate sustainable ecosystem management. Through the development of ecosystem monitoring, analysis, and management methods based on data science and artificial intelligence, the aim is to ensure Sustainability, Mitigation, Adaptation, Resilience to Global Change, and Trade-offs of Ecosystem Services (SMART) globally. Promoters, professors and researchers at the University of Valladolid (campus at Palencia) and the data and artificial intelligence team from SNGULAR, have been collaborating for twenty years in the development of programs and algorithms for ecosystem management. Through different collaborative projects developed with students from master DATAFOREST (University of Valladolid) and from Vietnam National University-University of Science (VNU University of Science), we have generated a set of lessons learned on the detection of singular trees by using ground and remote sensing data and artificial intelligence. Keywords: Innovation, Partnerships, Monitoring and data collection, Research, Learning by doing. ID: 3623999 -
ArticleWhy is artificial afforestation crucial for restoring nature? Studies on the dried bottom of Aral Sea, Kazakhstan
XV World Forestry Congress, 2-6 May 2022
2022Also available in:
No results found.The Aral Sea, formerly the fourth largest inland lake located in Central Asia, has reduced dramatically as it lost most of its volume due to the large-scale water withdrawal for the cultivation of irrigated crops starting from the 1960s. The desiccated seafloor has become a source of salt, sand, and dust transfer to the adjacent regions, negatively affecting human health and the environment by inhibiting the survival and growth of the vegetation. In response, to stabilize the saline sand blowing from the Aral seabed, multiple domestic and international efforts have been performed to establish vegetation cover with indigenous trees of Haloxylon species as well as other salt- and water-stress tolerant woody and herbaceous plants of the region. As part of the afforestation project supported by the Korea Forest Service (South Korea) in Kazakhstan, field studies examined the impacts of the afforestation on carbon stock and soil quality. The summarized findings are as follows: 1) growth of planted seedlings indicates the measurable sequestration of carbon, which ultimately help to estimate its contribution to climate change mitigation by calculating the atmospheric greenhouse gas reductions; 2) afforestation increased the soil organic matter content which is closely related to soil fertility; 3) afforestation improved soil chemical properties for plants and soil microbes; 4) soil amelioration effects by the afforestation were statistically similar to those by natural vegetation succession. However, the soil conditions in the natural succession area improved after almost 50 years versus about 15 years in the afforested area. This signifies the relative efficiency of afforestation activities and thus helps justify the investments made. Based on the studies, we recommend further research to raise the efficiency of afforestation in arid areas, thereby reinforcing ecosystem restoration and climate change mitigation. Keywords: Aral Sea; afforestation; desertification; climate change; restoration; soil amelioration; carbon stock ID: 3615605
Users also downloaded
Showing related downloaded files
No results found.