| Resumen |
The problem of retrieving images from a database is considered. In particular, we retrieve images belonging to
one of the following six categories: 1) commercial planes in land, 2) commercial planes in air, 3) war planes in
land, 4) war planes in air, 5) small aircraft in land, and 6) small aircraft in the air. During training, a wavelet-based
description of each image is first calculated using Daubechies 4-wavelet transformation. The resulting coefficients
are used to train a neural network (NN). During classification, test images are treated by the already trained NN.
Three different ways to obtain the coefficients of the Daubechies transform were proposed and tested: from the
entire image color channels, from the histogram of the biggest circular window inside the image color channels,
and from the histograms of the square sub-images in the image color channels of the original image. 120 images
were used for training and 240 for testing. The best efficiency of 88% was obtained with the third method.
Key Words: Image Retrieval, Wavelet Transform, Neural Classification.
|
