3-D texture model of natural scene image

Yoonsik Choe, R. L. Kashyap

Research output: Contribution to journalConference article

Abstract

A novel 3-D texture model is developed by considering the scene image as the superposition of a random texture image and a smooth shaded image. This smoothed surface is obtained by taking the average of pixel intensities from the natural image. A texture image can be generated by subtracting the smoothed image from the original one. Then, the shape-from-shading technique is applied to estimate the orientation of the surface. For describing a random texture image, the long-correlation periodic model is chosen because of its good performance in texture synthesis and its ability to represent the coarseness and the pattern of the surface at the same time. Then, a projection technique is developed to deal with a random model which has a nonisotropically distributed texture pattern. For estimating the parameters, a hybrid method which uses both the least-square and the maximum-likelihood estimates is applied directly to the given intensity function.

Original languageEnglish
Pages (from-to)463-467
Number of pages5
JournalConference Record - Asilomar Conference on Circuits, Systems & Computers
Volume1
Publication statusPublished - 1988 Dec 1
Eventv 1 (of 2) - Pacific Grove, CA, USA
Duration: 1988 Oct 311988 Nov 2

Fingerprint

Textures
Maximum likelihood
Pixels

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

@article{3b8b4113987f4b7388cb8ad813d3eb08,
title = "3-D texture model of natural scene image",
abstract = "A novel 3-D texture model is developed by considering the scene image as the superposition of a random texture image and a smooth shaded image. This smoothed surface is obtained by taking the average of pixel intensities from the natural image. A texture image can be generated by subtracting the smoothed image from the original one. Then, the shape-from-shading technique is applied to estimate the orientation of the surface. For describing a random texture image, the long-correlation periodic model is chosen because of its good performance in texture synthesis and its ability to represent the coarseness and the pattern of the surface at the same time. Then, a projection technique is developed to deal with a random model which has a nonisotropically distributed texture pattern. For estimating the parameters, a hybrid method which uses both the least-square and the maximum-likelihood estimates is applied directly to the given intensity function.",
author = "Yoonsik Choe and Kashyap, {R. L.}",
year = "1988",
month = "12",
day = "1",
language = "English",
volume = "1",
pages = "463--467",
journal = "Conference Record - Asilomar Conference on Circuits, Systems & Computers",
issn = "0736-5861",

}

3-D texture model of natural scene image. / Choe, Yoonsik; Kashyap, R. L.

In: Conference Record - Asilomar Conference on Circuits, Systems & Computers, Vol. 1, 01.12.1988, p. 463-467.

Research output: Contribution to journalConference article

TY - JOUR

T1 - 3-D texture model of natural scene image

AU - Choe, Yoonsik

AU - Kashyap, R. L.

PY - 1988/12/1

Y1 - 1988/12/1

N2 - A novel 3-D texture model is developed by considering the scene image as the superposition of a random texture image and a smooth shaded image. This smoothed surface is obtained by taking the average of pixel intensities from the natural image. A texture image can be generated by subtracting the smoothed image from the original one. Then, the shape-from-shading technique is applied to estimate the orientation of the surface. For describing a random texture image, the long-correlation periodic model is chosen because of its good performance in texture synthesis and its ability to represent the coarseness and the pattern of the surface at the same time. Then, a projection technique is developed to deal with a random model which has a nonisotropically distributed texture pattern. For estimating the parameters, a hybrid method which uses both the least-square and the maximum-likelihood estimates is applied directly to the given intensity function.

AB - A novel 3-D texture model is developed by considering the scene image as the superposition of a random texture image and a smooth shaded image. This smoothed surface is obtained by taking the average of pixel intensities from the natural image. A texture image can be generated by subtracting the smoothed image from the original one. Then, the shape-from-shading technique is applied to estimate the orientation of the surface. For describing a random texture image, the long-correlation periodic model is chosen because of its good performance in texture synthesis and its ability to represent the coarseness and the pattern of the surface at the same time. Then, a projection technique is developed to deal with a random model which has a nonisotropically distributed texture pattern. For estimating the parameters, a hybrid method which uses both the least-square and the maximum-likelihood estimates is applied directly to the given intensity function.

UR - http://www.scopus.com/inward/record.url?scp=0024182806&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0024182806&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:0024182806

VL - 1

SP - 463

EP - 467

JO - Conference Record - Asilomar Conference on Circuits, Systems & Computers

JF - Conference Record - Asilomar Conference on Circuits, Systems & Computers

SN - 0736-5861

ER -