AUTOMATED ROAD SEGMENT EXTRACTION BY GROUPING ROAD OBJECTS
A. P. Dal Poz * *, G. M. do Vale *, I. , R. B. Zanin*
“ Dept. of Cartography, Säo Paulo State University, Rua Roberto Simonsen, 305, 19060-900 Presidente Prudente, SP
$
Brazil - (aluir, gmvale, zanin)@prudente.unesp.br
Commission III, WG III/4
KEY WORDS: Photogrammetry, Vision, Automation, Recognition, Extraction, Edge, Object
ABSTRACT:
This article presents an automatic methodology for extraction of road segments from high-resolution aerial images. The method is
based on a set of four road objects and another set of connection rules among road objects. Each road object is a local representation
of an approximately straight road fragment and its construction is based on combination of polygons describing all relevant image
edges, according to some rules embodying road knowledge. Each road segments is composed by a sequence of connected road
objects, being each sequence of this type can be geometrically structured as a chain of contiguous quadrilaterals. Experiments
carried out with high-resolution aerial images showed that the proposed methodology is very promising for extracting road
segments. This article presents the fundamentals of the method, and the experimental results as well.
1. INTRODUCTION
Road extraction is of fundamental importance in context of
spatial data capturing and updating for GIS (Geographic
Information Systems) applications. Substantial work on road
extraction has been accomplished since the 70's in computer
vision and digital photogrammetry, with pioneering works by,
c. g., Bajcsy and Tavakoli (1976) and Quam (1978). At times
the use of term 'extraction' is vague, invoking varied meaning
among a diverse image analysis community. In this context, the
task of road extraction is related to two subtasks, i.e.:
recognition and delineation. By convention, road extraction
algorithm is categorised according to the extend to which it
addresses either subtask, thereby implying the relative level of
automation (Doucette et al., 2001). Usually, road extraction
methods that in principle do not need human interaction is
categorized as automatic, and the opposite as semi-automatic.
Thus, automatic methods address both road extraction subtasks
and semi-automatic methods address only the geometric
delineation of the roads, leaving the high-level decisions (i.e.,
the recognition) for a human operator, who uses his natural skill
to set the meaning to the object 'road'.
Concerning fully automatic methods, two basic steps can be
identified. The first one is the road segment extraction, in which
the local road properties tested are geometric (e.g.: roads are
smooth) and radiometric (e.g.: roads are usually lighter than the
background) in sense. As a result, only road segments or a
fragmented road network can be extracted. The second phase is
the road network completion, which requires a skilful
integration of contextual information (i.e., relations between
roads and other objects like trees and buildings) and other a
priori road knowledge into the road extraction methodology
(Baumgartner et al., 1999).
This paper only addresses the first phase of process for fully
automatic road network extraction. The motivation is the
fundamental importance of the road segment extraction for the
subsequent phase, as the potential success of this last phase is
significantly determined by the quality of the results of first
* Corresponding author.
phase. This paper is organised in four sections. Section 2
presents the proposed methodology for automatic road segment
extraction, which is essentially based on radiometric and
geometric road constraints. Preliminary results are presented
and discussed in Section 3. Conclusion and future perspectives
are provided in Section 4.
2. METHODOLOGY FOR AUTOMATIC ROAD
SEGMENT EXTRACTION
We propose a methodology for road segment extraction that is
based on a set of four road objects. Each road object is a local
representation of an approximately straight road fragment. The
road objects are sequentially connected to each other according
to a rule set, allowing road segments to be formed.
In the following, the extraction of road objects and the way they
are combined to construct road segments, are described with
enough details.
2.1 Extraction of Road Objects
The road objects are defined using straight line segments
belonging to two different polygons with characteristics that are
compatible to a road.
(a) Case 1
(c) Case 3
(d) Case 4
Figure 1. Road objects
Figure 1 shows the four road objects found in any road
segment. In the building of a road object, by convention the
inferior straight line segment is called base and the superior one
Internat
is callec
both s
orthogo
two pol
figure 1
are proj
The op]
cases 3
1(d), on
betweer
versa. Ii
candida.
combine
rectangl
each on
axis of ¢
The bui
construc
to identi
|. Anti-
gradi
belon
appro
appro
that 1
image
straig
Thus,
gradi
the m
2. Paral
line s
objec
close
3. Homt
much.
inside
homo
4. Conti
backg
its bac
. Super
segme
be ort
that g
figure
candi
onto
quadr
6. Fragr
polyg
relate:
short
very s
crossi
Anoth
shado
many
polyg
forme
cases
consic
Crossi
Strate
wn
