See veebileht kasutab küpsiseid kasutaja sessiooni andmete hoidmiseks. Veebilehe kasutamisega nõustute ETISe kasutustingimustega. Loe rohkem
Olen nõus
"Muu" projekt VLOTI15115
VLOTI15115 "Automatic drawing (1.10.2015−30.04.2016)", Gholamreza Anbarjafari, Tartu Ülikool, Loodus- ja tehnoloogiateaduskond, Tartu Ülikooli Tehnoloogiainstituut, Tartu Ülikool, Loodus- ja täppisteaduste valdkond, tehnoloogiainstituut.
Automatic drawing
Automatic drawing
Teadus- ja arendusprojekt
ValdkondAlamvaldkondCERCS erialaFrascati Manual’i erialaProtsent
4. Loodusteadused ja tehnika4.12. Protsessitehnoloogia ja materjaliteadusT111 Pilditehnika2.3. Teised tehnika- ja inseneriteadused (keemiatehnika, lennundustehnika, mehaanika, metallurgia, materjaliteadus ning teised seotud erialad: puidutehnoloogia, geodeesia, tööstuskeemia, toiduainete tehnoloogia, süsteemianalüüs, metallurgia, mäendus, tekstiilitehnoloogia ja teised seotud teadused).100,0
Super Appli, Inc.
01.10.2015−30.04.201636 312,50 EUR
36 312,50 EUR

Automatic drawing
Automatic drawing is a very challenging field of image processing and computer graphics due to endless possibilities of attributes of scenes. In this work we are planning to develop a system which is benefiting from the state-of-the-art and beyond the state-of-the-art techniques for creating automatic anime pictures. For this purpose, we collect the various anime cartoons and analyse the model of the cartoon as well as extracting and labelling different attributes of the design. Number of attributes will be indicated by Super Appli company’s artists. Based on the results of analysis, we present a new rendering method to represent the anime according to the chosen attribute by the artist. The selected attributes will help the system to show appropriate (in shape, in colour, and in properties) selection of the samples, so that the artist can choose the desired object. In order to conduct this automatic selection we will adopt support vector machine. In phase one of the project, some of the objects will have the possibility of manual modification. We will use the Lambertian shading model to determine the degree of the bright and dark colour of each vertex using one dimensional texture. Also we will adopt the cartoon rendering method using X-Toon method. Also we will use our recently introduced retexturing technique for enhancing the customization property of the automatic drawing system. The created drawing will be multi-layer, which lets the artist to manipulate each of the attributes independent from the other one and in order to obtain the final drawing merging and transformation (such as zooming) will be used. This results into having a dynamic output. The position of generated character can be determined by the artist, so the character will not necessarily be placed at any fixed position. Indeed its position depends on the background and the scenario that the artist has in his/her mind. In the first phase of the project the less detailed character and equipment will be generated and background information will be ignored. The system in phase one will also be able to keep track of the selection of the artists and any new modification of a chosen attribute will be keep as a new example of that attribute which results in more desirable drawing as the time pass.
Katse- ja arendustöö30,0