Primerjava različnih načinov določanja lesnatih rastlin

diplomsko delo - univerzitetni študij

Špela Trampuž (Author), Robert Brus (Mentor)

Abstract

V diplomski nalogi smo analizirali šest načinov določanja lesnatih rastlin, ki se razlikujejo med seboj po načinu določanja, zahtevnosti uporabe, dostopnosti, ciljni skupini ter obsegu vključenih taksonov. Preizkusili smo jih ob pomoči določanja petih drevesnih vrst, ki smo jih izbrali tako, da se med seboj bistveno razlikujejo po obliki in členjenosti listne ploskve, dve vrsti pa sta si močno podobni. Potek določanja smo podrobno beležili in med seboj primerjali tiste načine, pri katerih je bil proces beleženja enak. Med načini, pri katerih smo določanje lahko ocenili s številom korakov, potrebnih za določitev vrste, je najmanj korakov potreboval Interaktivni vodnik s kompleksnim dostopanjem, sledili so Ključ TNP, MFS in nazadnje Interaktivni vodnik z dihotomnim ključem. Število pravilno določenih vzorcev od skupno 19 je imela računalniška Aplikacija NS 13, mobilna aplikacija Leafsnap UK 11, ob ponovnem določanju pa 14. Med načini, namenjenimi splošni javnosti, se je v splošnem najbolj izkazal Ključ TNP. Med načini, ki zahtevajo strokovno znanja, je bil najbolj natančen ključ v Mali flori Slovenije, a je postopek določanja zelo zahteven. Aplikacija NS je veliko preprostejša za uporabo in se je pokazala za zanesljivo pri vrstah z značilno obliko lista, pri drugih vrstah pa rezultati niso bili dovolj zanesljivi. Enako težavo je imel Leafsnap UK. Nadaljnji razvoj sodobnih aplikacij za prepoznavanje lesnatih rastlin bo šel verjetno v dve smeri: po eni strani v razvoj algoritma za znanstveno zanesljivo prepoznavanje na osnovi skeniranih listov in/ali drugih rastlinskih delov, po drugi strani pa v razvoj široko uporabnega postopka za prepoznavanje vsaj osnovnega nabora vrst na osnovi fotografij listov na živih rastlinah.

Keywords

dendrologija;določanje lesnatih rastlin;določevalni ključi;lesnate rastline;načini določanja;

Data

Language:	Slovenian
Year of publishing:	2016
Typology:	2.11 - Undergraduate Thesis
Organization:	UL BF - Biotechnical Faculty
Publisher:	[Š. Trampuž]
UDC:	630*17(043.2)=163.6
COBISS:	4364454
Views:	454
Downloads:	110
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Comparison of different methods of woody plants determination
Secondary abstract:	In graduation thesis we analyzed six methods of determination for woody plants, which differ from each other in the manner of determining, the complexity of use, accessibility, target groups and the extent of included taxa. We tested them with determination of five tree species, which vary significantly in shape and composition of the leaf lamina, but two species are quite similar. Progress of the determination is recorded in detail. Then we compared methods, which had the same process of recording. Among methods, in which we can assess the determination in numbers of steps required to determine species, Interactive guide with a complex access needed the least steps, followed by the Key TNP, MFS and finally an Interactive guide with dichotomous key. Number of properly determined samples, from a total of 19, computer Application NS had 13, mobile application Leafsnap UK at first 11, after re-determining 14. Among the methods, aimed at the general public, Key TNP proved to be most useful. Among the ways that require professional knowledge, the key in Mala Flora Slovenije proves to be the most accurate, but the process of determining is very complex. Application NS is much easier to use and has been proved to be reliable for species with a characteristic shape of the leaf, at other species the results were not enough reliable. The same problem had Leafsnap UK. Further development of advanced applications for the identification of woody plants will go probably in two directions: On the one hand, the development of an algorithm for scientifically reliable identification on the basis of scanned sheets and / or other plant parts, on the other hand, the development of the widely used method for detecting at least a basic set species based on photographs of leaves on living plants.
Type (COBISS):	Undergraduate thesis
Thesis comment:	Univ. v Ljubljani, Biotehniška fakulteta, Oddelek za gozdarstvo in obnovljive gozdne vire
Pages:	X, 52 f., [11] f. barvnih pril.
ID:	11539742