Search results for "Contextual image classification"
showing 10 items of 105 documents
Towards a Hierarchical Multitask Classification Framework for Cultural Heritage
2018
Digital technologies such as 3D imaging, data analytics and computer vision opened the door to a large set of applications in cultural heritage. Digital acquisition of a cultural assets takes nowadays a couple of seconds thanks to the achievements in 2D and 3D acquisition technologies. However, enriching these cultural assets with labels and relevant metadata is still not fully automatized especially due to their nature and specificities. With the recent publication of several cultural heritage datasets, many researchers are tackling the challenge of effectively classifying and annotating digital heritage. The challenges that are often addressed are related to visual recognition and image c…
Kernel-Based Framework for Multitemporal and Multisource Remote Sensing Data Classification and Change Detection
2008
The multitemporal classification of remote sensing images is a challenging problem, in which the efficient combination of different sources of information (e.g., temporal, contextual, or multisensor) can improve the results. In this paper, we present a general framework based on kernel methods for the integration of heterogeneous sources of information. Using the theoretical principles in this framework, three main contributions are presented. First, a novel family of kernel-based methods for multitemporal classification of remote sensing images is presented. The second contribution is the development of nonlinear kernel classifiers for the well-known difference and ratioing change detectio…
Structured Output SVM for Remote Sensing Image Classification
2011
Traditional kernel classifiers assume independence among the classification outputs. As a consequence, each misclassification receives the same weight in the loss function. Moreover, the kernel function only takes into account the similarity between input values and ignores possible relationships between the classes to be predicted. These assumptions are not consistent for most of real-life problems. In the particular case of remote sensing data, this is not a good assumption either. Segmentation of images acquired by airborne or satellite sensors is a very active field of research in which one tries to classify a pixel into a predefined set of classes of interest (e.g. water, grass, trees,…
A multi-process system for HEp-2 cells classification based on SVM
2016
An automatic system for pre-segmented IIF images analysis was developed.A non-standard pipeline for supervised image classification was adopted.The system uses a two-level pyramid to retain some spatial information.From each cell image 216 features are extracted.15 SVM classifiers one-against-one have been implemented. This study addresses the classification problem of the HEp-2 cells using indirect immunofluorescence (IIF) image analysis, which can indicate the presence of autoimmune diseases by finding antibodies in the patient serum. Recently, studies have shown that it is possible to identify the cell patterns using IIF image analysis and machine learning techniques. In this paper we de…
Toward morphological thoracic EIT: major signal sources correspond to respective organ locations in CT.
2012
Lung and cardiovascular monitoring applications of electrical impedance tomography (EIT) require localization of relevant functional structures or organs of interest within the reconstructed images. We describe an algorithm for automatic detection of heart and lung regions in a time series of EIT images. Using EIT reconstruction based on anatomical models, candidate regions are identified in the frequency domain and image-based classification techniques applied. The algorithm was validated on a set of simultaneously recorded EIT and CT data in pigs. In all cases, identified regions in EIT images corresponded to those manually segmented in the matched CT image. Results demonstrate the abilit…
Interactive Pansharpening and Active Classification in Remote Sensing
2013
This chapter presents two multimodal prototypes for remote sensing image classification where user interaction is an important part of the system. The first one applies pansharpening techniques to fuse a panchromatic image and a multispectral image of the same scene to obtain a high resolution (HR) multispectral image. Once the HR image has been classified the user can interact with the system to select a class of interest. The pansharpening parameters are then modified to increase the system accuracy for the selected class without deteriorating the performance of the classifier on the other classes. The second prototype utilizes Bayesian modeling and inference to implement active learning …
Deep CNN-ELM Hybrid Models for Fire Detection in Images
2018
In this paper, we propose a hybrid model consisting of a Deep Convolutional feature extractor followed by a fast and accurate classifier, the Extreme Learning Machine, for the purpose of fire detection in images. The reason behind using such a model is that Deep CNNs used for image classification take a very long time to train. Even with pre-trained models, the fully connected layers need to be trained with backpropagation, which can be very slow. In contrast, we propose to employ the Extreme Learning Machine (ELM) as the final classifier trained on pre-trained Deep CNN feature extractor. We apply this hybrid model on the problem of fire detection in images. We use state of the art Deep CNN…
Support Vector Machines for Crop Classification Using Hyperspectral Data
2003
In this communication, we propose the use of Support Vector Machines (SVM) for crop classification using hyperspectral images. SVM are benchmarked to well–known neural networks such as multilayer perceptrons (MLP), Radial Basis Functions (RBF) and Co-Active Neural Fuzzy Inference Systems (CANFIS). Models are analyzed in terms of efficiency and robustness, which is tested according to their suitability to real–time working conditions whenever a preprocessing stage is not possible. This can be simulated by considering models with and without a preprocessing stage. Four scenarios (128, 6, 3 and 2 bands) are thus evaluated. Several conclusions are drawn: (1) SVM yield better outcomes than neura…
The indexing of persons in news sequences using audio-visual data
2004
We describe a video indexing system that automatically searches for a specific person in a news sequence. The proposed approach combines audio and video confidence values extracted from speaker and face recognition analysis. The system also incorporates a shot selection module that seeks for anchors, where the person on the scene is likely speaking. The system has been extensively tested on several news sequences with very good recognition rates.
Kernels for Remote Sensing Image Classification
2015
Classification of images acquired by airborne and satellite sensors is a very challenging problem. These remotely sensed images usually acquire information from the scene at different wavelengths or spectral channels. The main problems involved are related to the high dimensionality of the data to be classified and the very few existing labeled samples, the diverse noise sources involved in the acquisition process, the intrinsic nonlinearity and non-Gaussianity of the data distribution in feature spaces, and the high computational cost involved to process big data cubes in near real time. The framework of statistical learning in general, and of kernel methods in particular, has gained popul…