Growth estimation of standing trees using artificial intelligence
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
| Main Author: | |
|---|---|
| Format: | Document biblioteca |
| Language: | English |
| Published: |
FAO ;
2022
|
| Online Access: | https://openknowledge.fao.org/handle/20.500.14283/CC4429EN http://www.fao.org/3/cc4429en/cc4429en.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|