Extending the snake image processing algorithm for outlining pigs in scenes

Images containing animals are often difficult to process automatically because poor and variable lighting, shadows, and overlapping and touching animals lead to indistinct or incomplete boundaries. A snake is an algorithm based on the mechanics of an elastic loop. It can find continuous boundaries w...

Full description

Saved in:
Bibliographic Details
Published inComputers and electronics in agriculture Vol. 8; no. 4; pp. 261 - 275
Main Authors Marchant, J.A., Schofield, C.P.
Format Journal Article
LanguageEnglish
Published Amsterdam Elsevier B.V 1993
Elsevier
Subjects
algorithms
Animal productions
Biological and medical sciences
Fundamental and applied biological sciences. Psychology
image processors
pens
swine
Terrestrial animal productions
Vertebrates
Contourography
Overlapping
Farming animal
Image processing
Mechanical model
Edge boundary
Shadow
Algorithm
Pig
Image analysis
Vertebrata
Mammalia
Scene analysis
Lighting
Elastic strip
Artiodactyla
Computer aid
Ungulata
Online AccessGet full text

Cover

More Information
Summary:Images containing animals are often difficult to process automatically because poor and variable lighting, shadows, and overlapping and touching animals lead to indistinct or incomplete boundaries. A snake is an algorithm based on the mechanics of an elastic loop. It can find continuous boundaries when edge data suffer from the above problems. This paper briefly introduces the algorithm and then develops some novel extensions, specially useful for the problem of outlining pigs in a pen. The algorithm is tested on a number of images and shown to give good boundary location in cases where simpler methods fail.
ISSN:0168-1699
1872-7107
DOI:10.1016/0168-1699(93)90015-S