MISF: A Method for Measurement of Standing Tree Size via Multi-Vision Image Segmentation and Coordinate Fusion

With the development of computer vision technology, its applications in forestry are steadily becoming wider. To address problems inconvenience transporting unmanned aerial vehicles (UAVs), as well complex operation large instruments for measurement, a new method based on multi-vision image segmentation and coordinate fusion (MISF) is proposed this paper measuring size standing trees. In MISF, after images tree captured using camera from multiple angles, semantic deep learning used to segment main body automatically detect edge feature points. Next, effects visual field splicing analyzed collaboratively correlations among images, so restore three-dimensional spatial information points be measured. Lastly, attributes tree, such height, diameter at breast height (DBH), crown width, The urban environment measurement experiment showed that relative errors DBH, width measured method, i.e., were 1.89%, 2.42%, 3.15%, respectively, representing significant enhancement compared with binocular measurement. On one hand, experimental results exhibited high degree accuracy; therefore, MISF can management inventory typical forests. other cannot if tree’s acquired due environmental or reasons.