Publication detail

Object detection in color images based on neural networks

HYNČICA, T. JIRSÍK, V.

Original Title

Object detection in color images based on neural networks

Czech Title

Detekce objektů v barevných obrazech založený na neuronových sítích

English Title

Object detection in color images based on neural networks

Type

conference paper

Language

en

Original Abstract

Object detection is an important task in image processing and computer vision and artificial neural networks are very promising tool for such tasks. Since in most cases the color image contains much more information about image features than grayscale image, we focused on them. However, more complicated image processing methods are required to deal with processing of color images. The proper color representation (color space) is crucial for successful processing of color images. For that reason four color spaces (RGB, HSL, YCbCr and L*a*b*) are used. For object detection we used a RCE (Reduced Coulomb Energy) neural network.

Czech abstract

Detekce objektů je důležitým úkolem při zpracování obrazu a umělé neuronové sítě jsou velmi slibným nástrojem pro tento úkol. Vzhledem k tomu, že ve většině případů barevný obraz obsahuje mnohem více informací a obrazových příznaků než obraz ve stupních šedi, jsme se zaměřili na ně. Pro detekci objektů jsme použili RCE (Reduced Coulomb Energy) neuronové sítě.

English abstract

Object detection is an important task in image processing and computer vision and artificial neural networks are very promising tool for such tasks. Since in most cases the color image contains much more information about image features than grayscale image, we focused on them. However, more complicated image processing methods are required to deal with processing of color images. The proper color representation (color space) is crucial for successful processing of color images. For that reason four color spaces (RGB, HSL, YCbCr and L*a*b*) are used. For object detection we used a RCE (Reduced Coulomb Energy) neural network.

Keywords

RCE neural network, color spaces, object detection

RIV year

2012

Released

05.09.2012

Publisher

LITERA Brno

Location

Brno

ISBN

978-80-214-4547-5

Book

Workshop Perspektivní projekty vývoje řídicích a senzorických technologií září 2012

Edition number

1

Pages from

45

Pages to

49

Pages count

5

BibTex


@inproceedings{BUT93982,
  author="Tomáš {Hynčica} and Václav {Jirsík}",
  title="Object detection in color images based on neural networks",
  annote="Object detection is an important task in image processing and computer vision and artificial neural networks are very promising tool for such tasks. Since in most cases the color image contains much more information about image features than grayscale image, we focused on them. However, more complicated image processing methods are required to deal with processing of color images. The proper color representation (color space) is crucial for successful processing of color images. For that reason four color spaces (RGB, HSL, YCbCr and L*a*b*) are used. For object detection we used a RCE (Reduced Coulomb Energy) neural network.",
  address="LITERA Brno",
  booktitle="Workshop Perspektivní projekty vývoje řídicích a senzorických technologií září 2012",
  chapter="93982",
  howpublished="print",
  institution="LITERA Brno",
  year="2012",
  month="september",
  pages="45--49",
  publisher="LITERA Brno",
  type="conference paper"
}