DIGITALNA ARHIVA ŠUMARSKOG LISTA
prilagođeno pretraživanje po punom tekstu



ŠUMARSKI LIST 9-10/2008 str. 42     <-- 42 -->        PDF

D. Klobučar: PRIMJENA HISTOGRAMA DRUGOGA REDA U PROCJENI RELATIVNOG SASTOJINSKOG ... Šumarski list br. 9–10, CXXXII (2008), 419-429
L a n d e k , J., & D. K a u r i ć , 1998.: Ciklička aerofoto-litskih snimki. Diplomski rad, 65, Šumarski fagrametrijska
snimanja u Republici Hrvatskoj. kultet Zagreb.
Zbornik radova sto godina fotogrametrije u Hr-P r a n j i ć , A. & N. L u k i ć , 1997: Izmjera šuma. Za


tvatskoj. HAZU: 249–254, Zagreb. greb, 405 pp.
L i n d e r m a n , M., J. L i u , J. Q i , L. A n , Z. O u y -Quackenbush, L., P. Hopkins, & G. Kinn,


a n g s , J. Ya n g , & Y. Ta n , 2004: Using artifi2000:
Developing forestry products from high
cial neural networks to map the spatial distriburesolution
digital aerial imagery. Photogrammetion
of understorey bamboo from remote sentric
Engineering and Remote Sensing 66 (11),


sing data. Internationa Journal of Remote Senpp.
1337–1346.
sing, Vol. 25. No. 9, 1685–1700.


St-Onge, B., F. Cavayas, 1997: Automated forest
L o n č a r i ć , S., 2003: Predavanja. structure mapping from high resolution imagery
http://ipg.zesoi.fer.hr based on directional semivariogram estimates.
Pax-Lenney, M., C. E. Woodcock, S. A. Ma-Remote Sensing of Environment 61, pp. 82–95.
comber, S. Gopal, C. Song, 2001: Forest Verbeke, L. P. C., F. M. B. Va n Coillie, & R. R.
mapping with a generalized classifier and Land-D e Wu l f , 2006: Object-based forest stand densat
TM data. Remote Sensing of Environment 77, sity estimation from very high resolution optical
241–250. imagery using wavelet-based texture measures.


P e r n a r, R., & D. K l o b u č a r, 2003: Estimating stand In: 1st International Conference on Object-badensity
and condition with use of picture histosed
Image Analysis (OBIA 2006).
grams and visual interpretation of digital orthop-Wulder, M., K. Niemann, & D. Goodenough,
hotos. Glas. šum. pokuse 40: 81–111, Zagreb. 2000: Local maximum filtering for the extrac


Pernar, R., D. Klobučar, & V. Kušan, 2003: The tion of tree locations and basal area from high
application of aerial photographs from cyclic spatial resolution imagery. Remote Sensing of
recordings in the Republic of Croatia to forest Environment 73, pp. 103–114.


management. Glas. šum. pokuse 40: 113–168, Osnova gospodarenja za g. j . “Jamaričko brdo”, važ-
Zagreb. nost 1. 1. 2002. – 31. 12. 2011.
P o s a r i ć , D., 1996: Postupak dobivanja kolor kompo- Pravilnik o uređivanju šuma, NN 111/06.
zita kao podloga za vizualnu interpretaciju sate-


SUMMARY: The paper continues on past research into the application of
digital stand scene processing and digital ortophoto processing for the needs of
forest management. Black-and-white aerial photographs obtained during cyclic
surveying of the Republic of Croatia were used for this purpose at an approximate
scale 1:20,000 and 60 % overlap. Research was based on the example
of the management unit “Jamaričko Brdo” of Lipovljani Forest Office.


The most commonly used parameter for quantitative stand descriptions is
density. Since the application of digital processing techniques of stand scenes
in the last ten years has resulted in the development of new methods of stand
density evaluation, an additional possibility was investigated of cyclic photograph
application by constructing second order histograms and establishing
their relationship with three relative density categories.


First order histograms represent a graphic display of the proportion of some
numerical values in the pixels of a digital photograph. Shades of gray ranging
from 0 to 225 are found on the horizontal histogram axis, and the total
number of pixels with these shades is found on the vertical axis.


Texture measures calculated only by means of the first order histogram data
are deficient because they do not give information related to the relative relationship
between the pixels themselves.


Texture may be described by measuring the smudginess (width) of a histogram
around the main diagonal. Such type of textural data does not relate only
to intensity distribution (gray scale) but also to the position of pixels with
the same or similar values.