The tree-by-tree survey is an essential method for understanding forest species composition, structure, biomass, and dynamics, playing a crucial role in forest management, conservation, and planning. In a typical tree-by-tree survey, tree species, diameter at breast height (DBH), and tree height are measured and mapped. Measuring DBH requires physically reaching the target tree. Additionally, mapping is often conducted using a laser rangefinder, but branches and leaves of other trees can obstruct the laser, requiring adjustments in reference positions or the removal of obstacles with the help of others. According to the Forestry Agency's website, this method requires approximately 50 hours to survey 809 trees per hectare. Furthermore, additional time is needed to manually process and digitize the collected data. Due to the high level of expertise, labor, and time required, many forests remain unmonitored.
Recent smartphones are equipped with various sensors that enable easy acquisition of diverse data. Therefore, we have developed a program that allows anyone to conduct tree-by-tree surveys using only a smartphone. The program extracts location information from tree photos taken with a smartphone and measures DBH and tree height. The collected data is processed using a program developed in Python and visualized on a map. Additionally, we verified the accuracy of this method by comparing the obtained data with values measured using conventional methods.
Based on these results, we plan to implement augmented reality (AR) technology to display previously surveyed trees on the screen. Moreover, we are also considering developing a program that utilizes artificial intelligence (AI) to estimate tree species.