Search results for "Image processing"
showing 10 items of 3285 documents
Lensless object scanning holography for two-dimensional mirror-like and diffuse reflective objects
2013
Recently proposed lensless object scanning holography (LOSH) [Opt. Express 20, 9382 (2012)] is a fully lensless method capable of improving the image quality in digital Fourier holography applied to one-dimensional (1D) reflective objects and it involves a very simplified experimental setup. LOSH is based on the recording and digital postprocessing of a set of digital lensless Fourier transform holograms, which finally results in a synthetic image with improved resolution, field-of-view (FOV), signal-to-noise ratio (SNR), and depth of field. In this paper, LOSH is extended to the cases of two-dimensional (2D) mirror-like and 1D diffuse-based objects. For 2D mirror-like objects, the experime…
Computer-aided-diagnosis for ocular abnormalities from a single color fundus photography with deep learning
2023
Any damage to the retina can lead to severe consequences like blindness. This visual impairment is preventable by early detection of ocular abnormalities. Computer-aided diagnosis (CAD) for ocular abnormalities is built by analyzing retinal imaging modalities, for instance, Color Fundus Photography (CFP). The main objectives of this thesis are to build two CAD models, one to detect the microaneurysms (MAs), the first visible symptom of diabetic retinopathy, and the other for multi-label detection of 28 ocular abnormalities consisting of frequent and rare abnormalities from a single CFP by using deep learning-based approaches. Two methods were proposed for MAs detection: ensemble-based and c…
Reliable polygonal approximations of imaged real objects through dominant point detection
1998
Abstract The problem of dominant point detection is posed, taking into account what usually happens in practice. The algorithms found in the literature often prove their performance with laboratory contours, but the shapes in real images present noise, quantization, and high inter and intra-shape variability. These effects are analyzed and solutions to them are proposed. We will also focus on the conditions for an efficient (few points) and precise (low error) dominant point extraction that preserves the original shape. A measurement of the committed error (optimization error, E 0 ) that takes into account both aspects is defined for studying this feature.
Sequential Mining Classification
2017
Sequential pattern mining is a data mining technique that aims to extract and analyze frequent subsequences from sequences of events or items with time constraint. Sequence data mining was introduced in 1995 with the well-known Apriori algorithm. The algorithm studied the transactions through time, in order to extract frequent patterns from the sequences of products related to a customer. Later, this technique became useful in many applications: DNA researches, medical diagnosis and prevention, telecommunications, etc. GSP, SPAM, SPADE, PrefixSPan and other advanced algorithms followed. View the evolution of data mining techniques based on sequential data, this paper discusses the multiple …
Medical Data Mining for Heart Diseases and the Future of Sequential Mining in Medical Field
2018
Data Mining in general is the act of extracting interesting patterns and discovering non-trivial knowledge from a large amount of data. Medical data mining can be used to understand the events happened in the past, i.e. studying a patients vital signs to understand his complications and discover why he has died, or to predict the future by analyzing the events that had happened. In this chapter we are presenting an overview on studies that use data mining to predict heart failure and heart diseases classes. We will also focus on one of the trendiest data-mining field, namely the Sequential Mining, which is a very promising paradigm. Due to its important results in many fields, this chapter …
Overview on Sequential Mining Algorithms and Their Extensions
2018
The main purpose of data mining is to extract hidden, important and nontrivial information from a database. Sequential Pattern Mining is a data mining technique that aims to obtain and analyze frequent subsequences from sequences of events or items with or without time constraint. The importance of a sequence can be measured based on different factors such as the frequency of their occurrence, their length and also their profit. The pattern mining or the discovery of important and unexpected patterns and information was first introduced in 1990 with the well-known Apriori algorithm. Then, and after many studies on frequent pattern mining, a new approach appeared: Sequential Pattern Mining. …
Hop: Histogram of patterns for human action representation
2017
This paper presents a novel method for representing actions in terms of multinomial distributions of frequent sequential patterns of different length. Frequent sequential patterns are series of data descriptors that occur many times in the data. This paper proposes to learn a codebook of frequent sequential patterns by means of an apriori-like algorithm, and to represent an action with a Bag-of-Frequent-Sequential-Patterns approach. Preliminary experiments of the proposed method have been conducted for action classification on skeletal data. The method achieves state-of-the-art accuracy value in cross-subject validation.
Towards the Preservation and Dissemination of Historical Silk Weaving Techniques in the Digital Era
2019
Historical weaving techniques have evolved in time and space giving as result more or less fabrics with different aesthetical characteristics. These techniques were transferred along the main silk production centers, thanks to the European Silk Road and creating a common European Frame on themes and techniques. These had made it complicated to determine whether a fabric corresponds to one century or another. Moreover, in order to understand their creation, it is necessary to determine the number of weaves and interlacements that each textile has, therefore, mathematical models can be extracted from these layers. In this sense, three dimensional (3D) virtual representations of the internal s…
Statistical analysis of engraving traces on a 3D digital model of prehistoric stone stelae
2016
International audience; Studying cultural heritage artefacts, using 3D digital models, is gaining interest. It not only allows applications in documentation and visualisation, but also permits further contact-less examination. In this paper, we are presenting a statistical analysis of stone engravings based on features that were semi-automatically extracted from 3D acquisition data. Our objects of study are two Neolithic stone stelae and a faithful replica that was created in the course of an archaeological study. We use common statistical methods and investigate the populations of depth and diameter of the engraving traces, as well as their correlation. We observe that the erosion of the t…
Deep learning to detect built cultural heritage from satellite imagery. - Spatial distribution and size of vernacular houses in Sumba, Indonesia -
2021
Abstract In Sumba Island – Indonesia, the implantation of vernacular houses, inside and outside traditional villages, is considered to be an efficient proxy for the on-going complex cultural transformations resulting from globalization. This study presents an easily reproducible workflow allowing buildings to be automatically detected from satellite imagery, demonstrating how modern computer vision methods based on deep learning can help in this task, which would be far too time-consuming when undertaken by hand. Eight deep learning architectures based on convolutional neural networks were compared in terms of ability to identify and locate precisely traditional houses from satellite images…