Neural network based object recognition using color block matching

Authors:
Kay Boehnke;Marius Otesteanu;Philipp Roebrock;Wolfgang Winkler;Werner Neddermeyer
Affiliations:
Politehnica University of Timisoara, Timisoara, Romania;Politehnica University of Timisoara, Timisoara, Romania;Politehnica University of Timisoara, Timisoara, Romania;University of Applied Science, Gelsenkirchen, Germany;University of Applied Science, Gelsenkirchen, Germany
Venue:
SPPRA '07 Proceedings of the Fourth IASTED International Conference on Signal Processing, Pattern Recognition, and Applications
Year:
2007

Citing 4
Cited 0

Artificial intelligence: a modern approach

Artificial intelligence: a modern approach
Adaptive color reduction

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Digital color imaging

IEEE Transactions on Image Processing
Color clustering and learning for image segmentation based on neural networks

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper describes an algorithm for the fast classification of color regions in pictures with the help of neural networks. The algorithm divides the picture into discrete blocks which are analyzed independently. Average values of the three color channels are extracted from the block and classified by a neural network. Classification is made by a modified backpropagation network. After the classification a "best fit" search is used to find the best matching block. Depending on the block size it is possible to find and classify a learned color in a picture with the help of this sequential search algorithm. The proposed classification meets real time requirements in an industrial application. Therefore some optimizations were necessary which are explained in this paper. Furthermore, different pre-processing and segmentation algorithms and color space transformations were analyzed and tested regarding their effectiveness for fast image processing.