Publication detail

Real-Time Detection of Lines using Parallel Coordinates and CUDA

HAVEL, J. JURÁNKOVÁ, M. HEROUT, A. JOŠTH, R.

Original Title

Real-Time Detection of Lines using Parallel Coordinates and CUDA

English Title

Real-Time Detection of Lines using Parallel Coordinates and CUDA

Type

journal article in Web of Science

Language

en

Original Abstract

The Hough transform is a well-known and popular algorithm for detecting lines in raster images.  The standard Hough transform is rather slow to be usable in real-time, so different accelerated and approximated algorithms exist. This paper proposes a modified accumulation scheme for the Hough transform, using a new parameterization of lines "PClines". This algorithm is suitable for computer systems with a small but fast read-write memory -- such as today's graphics processors. The algorithm requires no floating-point computations or goniometric functions. This makes it suitable for special and low-power processors and special-purpose chips. The proposed algorithm is evaluated both on synthetic binary images and on complex real-world photos of high resolutions.  The results show that by using today's commodity graphics chips, the Hough transform can be computed at interactive frame rates, even with a high resolution of the Hough space and with the Hough transform fully computed.

English abstract

The Hough transform is a well-known and popular algorithm for detecting lines in raster images.  The standard Hough transform is rather slow to be usable in real-time, so different accelerated and approximated algorithms exist. This paper proposes a modified accumulation scheme for the Hough transform, using a new parameterization of lines "PClines". This algorithm is suitable for computer systems with a small but fast read-write memory -- such as today's graphics processors. The algorithm requires no floating-point computations or goniometric functions. This makes it suitable for special and low-power processors and special-purpose chips. The proposed algorithm is evaluated both on synthetic binary images and on complex real-world photos of high resolutions.  The results show that by using today's commodity graphics chips, the Hough transform can be computed at interactive frame rates, even with a high resolution of the Hough space and with the Hough transform fully computed.

Keywords

Hough Transform, PClines, CUDA, Real-Time Line Detection, Accumulation Scheme

RIV year

2012

Released

01.03.2014

Publisher

NEUVEDEN

Location

NEUVEDEN

ISBN

1861-8200

Periodical

Journal of Real-Time Image Processing

Year of study

2014

Number

9

State

DE

Pages from

205

Pages to

216

Pages count

12

Documents

BibTex


@article{BUT97062,
  author="Jiří {Havel} and Markéta {Juránková} and Adam {Herout} and Radovan {Jošth}",
  title="Real-Time Detection of Lines using Parallel Coordinates and CUDA",
  annote="The Hough transform is a well-known and popular algorithm for detecting lines in
raster images.  The standard Hough transform is rather slow to be usable in
real-time, so different accelerated and approximated algorithms exist.
This paper proposes a modified accumulation scheme for the Hough transform, using
a new parameterization of lines "PClines". This algorithm is suitable for
computer systems with a small but fast read-write memory -- such as today's
graphics processors. The algorithm requires no floating-point computations or
goniometric functions. This makes it suitable for special and low-power
processors and special-purpose chips. The proposed algorithm is evaluated both on
synthetic binary images and on complex real-world photos of high resolutions. 
The results show that by using today's commodity graphics chips, the Hough
transform can be computed at interactive frame rates, even with a high resolution
of the Hough space and with the Hough transform fully computed.",
  address="NEUVEDEN",
  chapter="97062",
  doi="10.1007/s11554-012-0303-4",
  edition="NEUVEDEN",
  howpublished="online",
  institution="NEUVEDEN",
  number="9",
  volume="2014",
  year="2014",
  month="march",
  pages="205--216",
  publisher="NEUVEDEN",
  type="journal article in Web of Science"
}