“Tree diameter at breast height measurement with smartphone sensors,“ (2026)
/Z. Yin, S. Fei and S. Zhang, “Tree diameter at breast height measurement with smartphone sensors,“ Optics and Lasers in Engineering, 201, 109684 (2026)
Abstract:
Diameter at breast height (DBH) is a key structural parameter in forestry, widely used in forest inventory, biomass estimation, and ecological modeling. This paper presents an automated DBH measurement method using a single snapshot from an iPhone’s LiDAR and red, green, blue (RGB) sensors. The proposed method integrates automatic breast height localization from a user-tapped base point, edge-guided segmentation of the trunk, and tree diameter estimation. The method was first validated on a light pole, achieving a mean absolute error (MAE) of 1.3 cm and a root-mean-square error (RMSE) of 1.6 cm across 75 samples with different distances and viewing angles. It was further evaluated on 202 planted trees, yielding a coefficient of determination R2 of 0.966, a MAE of 1.1 cm, a RMSE of 1.5 cm, MAE percentage of 3.2%, and RMSE percentage of 4.3%, which are sufficient for tree inventory. Additionally, the method was evaluated on 287 trees in a complex natural forest environment, where it initially yielded a coefficient of determination (R2) of -0.949 due to occlusions from branches or foliage. However, after excluding the six segmentation-induced outliers, it yields a coefficient of determination R2 of 0.841, with MAE and MAE% 2.1cm and 5.7%, indicating that once the segmentation failures are excluded, the geometric measurement model remains stable and accurate even in complex natural forest settings.