No abstract

No abstract

Abstract-Texture analysis plays an important role in computer vision and pattern recognition applications. During the last few decades, the research community has proposed a large number of techniques for describing, retrieving and classifying texture images. Local Binary Patterns (LBP) coding is a state-of-the-art technique characterized by its simplicity and efficiency. Due to its success, several LBP-variants are proposed in recent literature. In this paper we show that the performance of LBP-based methods can be further improved by introducing a simple modification to the feature extraction process. We suggest building two different LBP histograms one for edge pixels and the second for non-edge pixels. The final feature vector is a weighted combination of the two histograms. This idea is mainly inspired by the results of several research works on vision indicating that when looking at objects, human attention focuses more on salient regions (where changes in intensity, color, etc. occur). The experiments that have been conducted on Brodatz and Outex databases show that implementing this modification on LBP-based techniques (LBP, LTP and LBP_V), produces significant improvement in the accuracy of the original methods.Index Terms-Edge information, local binary patterns, texture analysis.

No abstract

Alzheimer’s disease (AD) is a neurodegenerative disease that causes irreversible damage to several brain regions, including the hippocampus causing impairment in cognition, function, and behaviour. Early diagnosis of the disease will reduce the suffering of the patients and their family members. Towards this aim, in this paper, we propose a Siamese Convolutional Neural Network (SCNN) architecture that employs the triplet-loss function for the representation of input MRI images as k-dimensional embeddings. We used both pre-trained and non-pretrained CNNs to transform images into the embedding space. These embeddings are subsequently used for the 4-way classification of Alzheimer’s disease. The model efficacy was tested using the ADNI and OASIS datasets which produced an accuracy of 91.83% and 93.85%, respectively. Furthermore, obtained results are compared with similar methods proposed in the literature.