In this work is presented a pattern recognition image descriptor invariant to rotation, scale and translation (RST), which classify images using the Z-Fisher transform. A binary rings mask is generated using the Fourier transform. The normalized analytic Fourier-Mellin amplitude spectrum is filtered with that mask to build 1D signature. The signatures comparison of the problem image and the target are done by the Pearson correlation coe cient (PCC). In general, those PCC values do not satisfy a normal distribution, hence the Fisher's Z distribution is employed to determine the confidence level of the RST invariant descriptor. The descriptor presents a confidence level of 95%.
This work presents a color image pattern recognition system invariant to rotation, scale and translation. The system works with three 1D signatures, one for each RGB color channel. The signatures are constructed based on Fourier transform, analytic Fourier-Mellin transform and Hilbert binary rings mask. According with the statistical theory of box-plots, the pattern recognition system has a confidence level at least of 95.4%.
In this paper is presented a pattern recognition methodology based on local feature extraction. The purpose of this system is to identify and locate, in three different scale pyramids, key points that represent relevant information of the image; this information is stored in a descriptor which is used to compares the key points of two images and know if they have similar information, or if they are the same images. This methodology uses the Haar wavelet transform to generate the three scale pyramids. This transform is used because it has several properties, such as noise elimination, multi-resolution analysis, and detection of diagonal, horizontal and vertical edges. The performance of this system was tasted using images with different scales and comparing the results with the Scale Invariant Feature Transform (SIFT) and Speeded-Up Robust Features (SURF) methodologies. The WLF system showed to has the highest percentage of correct point-matching.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.