Gonzalez, R.C., Woods, R. E., 2002: Digital Image Processing. Prentice Hall, Inc., Upper Saddle River, New Jersey, second edition, 793 p.
Hirschmugl, M., Gallaun, H., Wack, R., Granica, K., Schardt, M., 2013: EUFODOS: European Forest Downstream Services Improved Information on Forest Structure and Damage Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL1/W1: 127-131.
Maltamo, M., K. Eerikainen, J. Pitkanen, J. Hyyppa, and M. Vehmas., 2004. Estimation of timber volume and stem density based on scanning laser altimetry and expected tree size distribution functions. Remote Sensing of Environment. 90: 319-330. [OpenAIRE]
Schardt, M., Ziegler. M., Wimmer, A., Wack, R., Hyyppa, J. 2002: Assessment of forest parameters by mean of laser scanning. ISPRS Arch. Vol. 34(3), 304-312.
Schardt, M., Granca, K., Hirschmugl, M., Deutscher, J., Mollatz, M., Steinegger, M., Gallun, H., Wimmer, A., Linser, S., 2015: The assessment of forest parameters by combined LiDAR and satellite data over Alpine regions-EUFODO S Implementation in Austria Lesn. Cas. For. J. 61 (2015) 3-11. [OpenAIRE]
Sessions, J., A. E. Akay, G. Murphy, W. Chung, and K. Aruga., 2006. Road and Harvesting Planning and Operations in Computer Applications in Sustainable Forest Management, Including Perspectives on Collaboration and Integration Series: Managing Forest Ecosystems, Vol. 11, Edited by Shao, Guofan and Reynolds, Keith M. Sipringer, Berlin. 2006, XXXII, 277 p. ISBN:978-1-4020-4305-5.
Reutebuch, S.E., R.J. McGauhey, H.E. Andersen, and W.W. Carson., 2003. Accuracy of high-resolution LIDAR terrain model under a conifer forest canopy. Canadian J. of Remote Sensing. 29(5):527-535.
Wack, R., 2006. Assessment of Alpine Protection Forest Parameters based on LIDAR Data ad SPOT V Satellite Imagery. Proceedings of th HMRSC Workshop Graz, Austria.
Airborne LiDAR data have been collected for the city of Istanbul using Riegl laser scanner Q680i with 400 kHz and an average flight height of 600 m. The flight campaign was performed by a helicopter and covers an area of 5400 km2. According to a flight speed of 80 knot a point density of more than 16 points/m2 and a laser footprint size of 30 cm could be achieved. As a result of bundle adjustment, in total, approximately 17,000 LAS files with the file size of 500 m by 700 m have been generated for the whole city. The main object classes Ground, Building, Vegetation (medium, high) were derived from these LAS files using the macros in Terrasolid software. The forest area under investigation is located northwest of the city of Istanbul, main tree species occurring in the test site are pine (pinus pinaster), oak (quercus) and beech (fagus). In total, 120 LAS tiles covering the investigation area have been analysed using the software IMPACT of Joanneum Research Forschungsgesellschaft, Graz, Austria. First of all, the digital terrain model (DTM) and the digital surface models (DSM) were imported and converted into a raster file from the original laser point clouds with a spatial resolution of 50 cm. Then, a normalized digital surface model (nDSM) was derived as the difference between DSM and the DTM. Tree top detection was performed by multi – resolution filter operations and tree crowns were segmented by a region growing algorithms develop specifically for this purpose. Breast Height Diameter (BHD) was calculated on the base of tree height and crown areas derived from image segmentation applying allometric functions found in literature. The assessment of stem volume was then calculated as a function of tree height and BHD. A comparison of timber volume estimated from the LiDAR data and field plots measured by the Forest Department of Istanbul showed R2 of 0.46. The low correlation might arise either from the low quality of the field plots or from the inadequacy of the allometric functions used for BHD and stem volume modelling. Further investigations therefore will concentrate both on improving the quality of field measurements and the adoption of the allometric functions to the specific site condition of the forests under investigation. Finally stand boundaries of the forest area made available by the forest department of Istanbul were superimposed to the LiDAR data and the single tree measurements aggregated to the stand level. Aside from the LiDAR data, a Pleiades multispectral image characterized by four spectral bands (blue, green, red and near infrared) and a GSD of 2.8 m has been used for the classification of different tree species. For this purpose the near infrared band covering the spectral range of 0.75 μm to 0.90 μm has been utilized and the IMPACT software used.