Title :
Automatic tree stem delineation supporting forest inventory
Author :
Molinier, Matthieu ; Andersson, Karl ; Hame, Tuomas
Author_Institution :
Digital Inf. Syst., VTT Tech. Res. Centre of Finland, Espoo, Finland
Abstract :
The aim of this study was to develop an automatic procedure to detect tree stems in cell phone images. If the stems can be detected, the central projection geometry of the images makes it possible to apply the so called relascope principle to predict the basal area of stems per hectare. The basal area is closely correlated with tree biomass. Images were acquired in an area where detailed tree-wise information is available. The tree stem delineation methods were based on color and edge information to tackle the challenges of occlusions and varying illumination conditions. Geolocalised images can be used as in-situ reference data for remote sensing data analysis to provide wall-to-wall estimates.
Keywords :
edge detection; image colour analysis; mobile handsets; object detection; automatic tree stem delineation method; cell phone images; central projection geometry; color information; detailed tree-wise information; edge information; forest inventory; geolocalised images; illumination conditions; in-situ reference data; relascope principle; remote sensing data analysis; stem basal area; tree biomass; tree stem detection; Cellular phones; Forestry; Image color analysis; Image edge detection; Image segmentation; Lighting; Vegetation; Cellular phone image processing; Forestry; Object detection and tree recognition;
Conference_Titel :
Geoscience and Remote Sensing Symposium (IGARSS), 2011 IEEE International
Conference_Location :
Vancouver, BC
Print_ISBN :
978-1-4577-1003-2
DOI :
10.1109/IGARSS.2011.6129458