Learning vector quantization with alternative distance criteria

An adaptive algorithm for training of a nearest neighbour (NN) classifier is developed in this paper. This learning rule has some similarity to the well-known LVQ method, but uses the nearest centroid neighbourhood concept to estimate optimal locations of the codebook vectors. The aim of this approach is to improve the performance of the standard LVQ algorithms when using a very small codebook. The behaviour of the learning technique proposed here is experimentally compared to those of the plain k-NN decision rule and the LVQ algorithms.

Affine Illumination Compensation on Hyperspectral/Multiangular Remote Sensing Images

The huge amount of information some of the new optical satellites developed nowadays will create demands to quickly and reliably compensate for changes in the atmospheric transmittance and varying solar illumination conditions. In this paper three different forms of affine transformation models (general, particular and diagonal) are considered as candidates for rapid compensation of illumination variations. They are tested on a group of three pairs of CHRISPROBA radiance images obtained in a test field in Barrax (Spain), and where there is a difference in the atmospheric as well as in the geometrical acquisition conditions. Results indicate that the proposed methodology is satisfactory for …

Location and characterization of the stem-calyx area on oranges by computer vision

Three image analysis methods were studied and evaluated to solve the problem of removing long stems attached to mechanically harvested oranges: colour segmentation based on linear discriminant analysis, contour curvature analysis, and a thinning process which involves iterating until the stem becomes a skeleton. These techniques are able to determine the presence or absence of a stem with certainty, to locate the stems from random views with more than 90% accuracy and from profile images with an accuracy ranging from 92.4% to 100% depending on the method used. Finally, determination of the length and cutting point of the stem is achieved with only 3.8% of failures. (C) 1996 Silsoe Research …

From the nearest neighbour rule to decision trees

This paper proposes an algorithm to design a tree-like classifier whose result is equivalent to that achieved by the classical Nearest Neighbour rule. The procedure consists of a particular decomposition of a d-dimensional feature space into a set of convex regions with prototypes from just one class. Some experimental results over synthetic and real databases are provided in order to illustrate the applicability of the method.

Improving the k-NCN classification rule through heuristic modifications

Abstract This paper presents an empirical investigation of the recently proposed k-Nearest Centroid Neighbours ( k -NCN) classification rule along with two heuristic modifications of it. These alternatives make use of both proximity and geometrical distribution of the prototypes in the training set in order to estimate the class label of a given sample. The experimental results show that both alternatives give significantly better classification rates than the k -Nearest Neighbours rule, basically due to the properties of the plain k -NCN technique.

Roadmap on 3D integral imaging: Sensing, processing, and display

This Roadmap article on three-dimensional integral imaging provides an overview of some of the research activities in the field of integral imaging. The article discusses various aspects of the field including sensing of 3D scenes, processing of captured information, and 3D display and visualization of information. The paper consists of a series of 15 sections from the experts presenting various aspects of the field on sensing, processing, displays, augmented reality, microscopy, object recognition, and other applications. Each section represents the vision of its author to describe the progress, potential, vision, and challenging issues in this field.

Colour segmentation based on a light reflection model to locate citrus fruits for robotic harvesting

Abstract Colour segmentation with a vision system is a good procedure to identify and locate fruits in robotic harvesting. Natural illumination conditions present in these environments produce a very variable illumination of the scene, in addition, fruits are usually partially occluded, and complete visual information about them is not available. The colour segmentation used for these purposes must take into account the appearance of highlights and shadows that natural illumination conditions produce. A method based on the Dichromatic Reflection Model for the light reflected from the surface object is reported here. Through the assumption of this model the light rays reflected from points o…

Affine compensation of illumination in hyperspectral remote sensing images

A problem when working with optical satellite or airborne images is the need to compensate for changes in the illumination conditions at the time of acquisition. This is particularly critical when working with time series of data. Atmospheric correction strategies based on radiative transfer codes may provide a rigorous solution but it may not be the best solution for situations where a huge amount of hyperspectral images may need to be processed and computational time is a critical factor. The GMES (”Global Monitoring for Environment and Security”) initiative has promoted the creation of a new generation of satellites (the SENTINEL series) with ”ultra-high resolution” and ”superspectral im…

Sentinel-3/FLEX Biophysical Product Confidence Using Sentinel-2 Land-Cover Spatial Distributions

The estimation of biophysical variables from remote sensing data raises important challenges in terms of the acquisition technology and its limitations. In this way, some vegetation parameters, such as chlorophyll fluorescence, require sensors with a high spectral resolution that constrains the spatial resolution while significantly increasing the subpixel land-cover heterogeneity. Precisely, this spatial variability often makes that rather different canopy structures are aggregated together, which eventually generates important deviations in the corresponding parameter quantification. In the context of the Copernicus program (and other related Earth Explorer missions), this article propose…

Fundamentals of automated human gesture recognition using 3D integral imaging: a tutorial

Automated human gesture recognition is receiving significant research interest, with applications ranging from novel acquisition techniques to algorithms, data processing, and classification methodologies. This tutorial presents an overview of the fundamental components and basics of the current 3D optical image acquisition technologies for gesture recognition, including the most promising algorithms. Experimental results illustrate some examples of 3D integral imaging, which are compared to conventional 2D optical imaging. Examples of classifying human gestures under normal and degraded conditions, such as low illumination and the presence of partial occlusions, are provided. This tutorial…

Multidimensional optical sensing and imaging for displays, computational imaging, optical security, and healthcare

In this invited paper, we present an overview of our recently published work on 3D imaging, visualization and displays, including optical security using quantum imaging principles, 3D microscopy, healthcare, automated disease identification with 3D imaging, fatigue free augmented reality 3D glasses, and optical security and authentication using photon counting for IC inspection, polarimetric photon counting 3D imaging, and 3D human gesture recognition

Complexity reduction in efficient prototype-based classification

Multispectral integral imaging acquisition and processing using a monochrome camera and a liquid crystal tunable filter

This paper presents an acquisition system and a procedure to capture 3D scenes in different spectral bands. The acquisition system is formed by a monochrome camera, and a Liquid Crystal Tunable Filter (LCTF) that allows to acquire images at different spectral bands in the [480, 680]nm wavelength interval. The Synthetic Aperture Integral Imaging acquisition technique is used to obtain the elemental images for each wavelength. These elemental images are used to computationally obtain the reconstruction planes of the 3D scene at different depth planes. The 3D profile of the acquired scene is also obtained using a minimization of the variance of the contribution of the elemental images at each …

SmartSpectra: Applying multispectral imaging to industrial environments

SmartSpectra is a smart multispectral system for industrial, environmental, and commercial applications where the use of spectral information beyond the visible range is needed. The SmartSpectra system provides six spectral bands in the range 400-1000nm. The bands are configurable in terms of central wavelength and bandwidth by using electronic tunable filters. SmartSpectra consists of a multispectral sensor and the software that controls the system and simplifies the acquisition process. A first prototype called Autonomous Tunable Filter System is already available. This paper describes the SmartSpectra system, demonstrates its performance in the estimation of chlorophyll in plant leaves, …

Using proximity and spatial homogeneity in neighbourhood-based classifiers

In this paper, a set of neighbourhood-based classifiers are jointly used in order to select a more reliable neighbourhood of a given sample and take an appropriate decision about its class membership. The approaches introduced here make use of two concepts: proximity and symmetric placement of the samples.

On the use of neighbourhood-based non-parametric classifiers

Alternative non-parametric classification schemes, which come from the use of different definitions of neighbourhood, are introduced. In particular, the Nearest Centroid Neighbourhood along with the neighbourhood relation derived from the Gabriel Graph and the Relative Neighbourhood Graph are used to define the corresponding (k-)Nearest Neighbour-like classifiers. Experimental results are reported to compare the performance of the approaches proposed here to the one obtained with the k-Nearest Neighbours rule.

Multitemporal Mosaicing for Sentinel-3/FLEX Derived Level-2 Product Composites

The increasing availability of remote sensing data raises important challenges in terms of operational data provision and spatial coverage for conducting global studies and analyses. In this regard, existing multitemporal mosaicing techniques are generally limited to producing spectral image composites without considering the particular features of higher-level biophysical and other derived products, such as those provided by the Sentinel-3 (S3) and Fluorescence Explorer (FLEX) tandem missions. To relieve these limitations, this article proposes a novel multitemporal mosaicing algorithm specially designed for operational S3-derived products and also studies its applicability within the FLEX…

Multidimensional Optical Sensing and Imaging Systems (MOSIS): From Macro to Micro Scales

Multidimensional optical imaging systems for information processing and visualization technologies have numerous applications in fields such as manufacturing, medical sciences, entertainment, robotics, surveillance, and defense. Among different three-dimensional (3-D) imaging methods, integral imaging is a promising multiperspective sensing and display technique. Compared with other 3-D imaging techniques, integral imaging can capture a scene using an incoherent light source and generate real 3-D images for observation without any special viewing devices. This review paper describes passive multidimensional imaging systems combined with different integral imaging configurations. One example…

Editing prototypes in the finite sample size case using alternative neighborhoods

The recently introduced concept of Nearest Centroid Neighborhood is applied to discard outliers and prototypes 111 class overlapping regions in order to improve the performance of the Nearest Neighbor rule through an editing procedure, This approach is related to graph based editing algorithms which also define alternative neighborhoods in terms of geornetric relations, Classical editing algorithms are compared to these alternative editing schemes using several synthetic and real data problems. The empirical results show that, the proposed editing algorithm constitutes a good trade-off among performance and computational burden.

Multidimensional Integral Imaging for Sensing, Visualization, and Recognition in Degraded Environments

An overview of multidimensional integral imaging for sensing, visualization, and recognition in degraded environments is presented. Applications include 3D visualization, photon starved imaging, material inspection, IR imaging, passive depth estimation, automated human gesture recognition, and long-range imaging.

Multidimensional Integral Imaging and Recognition in Degraded Environments

We present an overview of our work on multidimensional integral imaging systems. Integral-imaging-based multidimensional optical sensing and imaging will be described for 3-D visualization, seeing through obscurations, material inspection, augmented reality, biomedical applications, and object recognition from microscales to long-range imaging.

Desarrollo de productos avanzados para la misión SEOSAT/Ingenio

SEOSAT/Ingenio es la futura misión española de observación de la Tierra en el óptico en alta resolución espacial. Mientras que los productos de imagen a Nivel 1, radiancias geo-referenciadas a nivel de sensor, se encuentran en una fase avanzada de desarrollo existiendo para ello un contrato industrial, los productos de Nivel 2 deben ser desarrollados por los propios usuarios. Este hecho limita el uso de las imágenes a la comunidad científica, restringiendo sus posibles aplicaciones fuera de ésta. Así pues, bajo el marco de un proyecto coordinado y motivados por ofrecer productos de Ingenio/SEOSAT de Nivel 2 a disposición de cualquier usuario, se origina y desarrolla este trabajo. En este ar…

Estimating feature discriminant power in decision tree classifiers

Feature Selection is an important phase in pattern recognition system design. Even though there are well established algorithms that are generally applicable, the requirement of using certain type of criteria for some practical problems makes most of the resulting methods highly inefficient. In this work, a method is proposed to rank a given set of features in the particular case of Decision Tree classifiers, using the same information generated while constructing the tree. The preliminary results obtained with both synthetic and real data confirm that the performance is comparable to that of sequential methods with much less computation.

Combining Defocus and Photoconsistency for Depth Map Estimation in 3D Integral Imaging

This paper presents the application of a depth estimation method for scenes acquired using a Synthetic Aperture Integral Imaging (SAII) technique. SAII is an autostereoscopic technique consisting of an array of cameras that acquires images from different perspectives. The depth estimation method combines a defocus and a correspondence measure. This approach obtains consistent results and shows noticeable improvement in the depth estimation as compared to a minimum variance minimisation strategy, also tested in our scenes. Further improvements are obtained for both methods when they are fed into a regularisation approach that takes into account the depth in the spatial neighbourhood of a pix…

Prototype selection for the nearest neighbour rule through proximity graphs

Abstract In this paper, the Gabriel and Relative Neighbourhood graphs are used to select a suitable subset of prototypes for the Nearest Neighbour rule. Experiments and results are reported showing the effectiveness of the method and comparing its performance to those obtained by classical techniques.