Conjoining Gestalt Rules for Abstraction of Architectural Drawings

晃晃

Conjoining Gestalt Rules for Abstraction of Architectural Drawings

Liangliang Nan1 Andrei Sharf2 Ke Xie1 Tien-Tsin Wong3 Oliver Deussen4 Daniel Cohen-Or

5 Baoquan Chen1

1 SIAT 2 Ben Gurion Univ.3 CUHK 4 Konstanz Univ.5 Tel Aviv Univ.









Abstract

We present a method for s***ctural summarization and abstraction

of complex spatial arrangements found in architectural drawings.

The method is based on the well-known Gestalt***les, which sum-

marize how forms, patterns, and semantics are perceived by humans

from bits and pieces of geometric information. Although defining

a computational model for each***le alone has been extensively s-

tudied, modeling a conjoint of Gestalt***les remains a challenge.

In this work, we develop a computational framework which mod-

els Gestalt***les and more importantly, their complex interaction-

s. We apply conjoining***les to line drawings, to detect groups of

objects and repetitions that conform to Gestalt principles. We sum-

marize and abstract such groups in ways that maintain s***ctural

semantics by displaying only a reduced number of repeated ele-

ments, or by replacing them with simpler shapes. We show an ap-

plication of our method to line drawings of architectural models of

various styles, and the potential of extending the technique to other

computer-generated illustrations, and three-dimensional models.

CR Categories: I.3.5 [Computer Graphics]: Geometric Modeling

and Level of Detail—Geometric algorithms;

Keywords: simplification, abstraction, Gestalt

1 Introduction

Artistic imagery, architectural renderings, cartography and games

often exploit abstraction to clarify, exaggerate, simplify or empha-

size the visual content. Abstraction is a strategy for communicating

information effectively. It allows artists to highlight specific visu-

al information and thereby direct the viewer to important aspects

of the s***cture and organization of the scene. In this paper, we

present an approach to the abstraction of 2D shapes, in particular

those of architectural models. Our approach to abstracting shape

directly aims to clarify shape and preserve meaningful s***ctures

using Gestalt principles.

The well-known Gestalt principles by Wertheimer [1923], reflect

strategies of the human visual system to group objects into forms

and create internal representations for them. Whenever groups of

visual element have one or several characteristics in common, they

get grouped and form a new larger visual object - a gestalt. Psychol-

ogists have tried to simulate and model these principles, by finding

computational means to predict what human perceive as gestalts in

images.

The notion of Gestalt is very well-known and widely used in var-

ious fields. In particular, it explains the tendency of the human

visual recognition to form whole shapes and forms just from bits

and pieces of geometric information. Naturally, Gestalt principles

have been used in computer vision, primarily in context with object

recognition and scene understanding. In computer graphics, Gestalt

principles have been applied to a variety of applications, like scene

completion [Drori et al. 2003], image and scene abstraction [Wang

et al. 2004; Mehra et al. 2009], stroke synthesis [Barla et al. 2006;

Ijiri et al. 2008] and emerging images generation [Mitra et al. 2009].

In general, these works rely on discrete Gestalt principles, but none

addresses the complex interactions emerging from the multitude of

Gestalt principles operating simultaneously.

A difficult problem while dealing with gestalts is the conjoined ef-

fect of two or more Gestalt principles operating at the same time

on the same site. Modeling gestalts in such cases is especially

challenging due to the complexity and ambiguity of the scene. Re-

cently, attempts to discover how grouping principles interact were

made in psychology and computer vision [Desolneux et al. 2002;

Feldman 2003; Cao et al. 2007; Kubovy and van den Berg 2008].

These works model limited gestalt interactions, by finding com-

putational means which are physiologically plausible. Kubovy and

van den Berg [2008] explore the quantification of perceptual group-

ings formed conjointly by two grouping principles: similarity and

proximity. Nevertheless, providing general computational means

for modeling the interaction of multiple Gestalt principles remains

a difficult challenge.









全文请下载附件：

tc

很经典，很实用，学习了!

晃晃

赞一个，哈哈

tc

楼主收集的可真全哦

奇

很有心,部分已收录自用,谢谢

奇

加精、加亮滴铁子，尤其要多丁页丁页

tc

跑着去顶朋友滴铁

tc

有意思！学习了！

菜刀吻电线

跑着去顶朋友滴铁

tc

很经典，很实用，学习了!