Examination of the quality of spinach leaves using hyperspectral imaging

The present research is focused on the application of hyperspectral images for the supervision of quality deterioration in ready to use leafy spinach during storage (Spinacia oleracea). Two sets of samples of packed leafy spinach were considered: (a) a first set of samples was stored at 20 °C (E-20) in order to accelerate the degradation process, and these samples were measured the day of reception in the laboratory and after 2 days of storage; (b) a second set of samples was kept at 10 °C (E-10), and the measurements were taken throughout storage, beginning the day of reception and repeating the acquisition of Images 3, 6 and 9 days later. Twenty leaves per test were analyzed. Hyperspectral images were acquired with a push-broom CCD camera equipped with a spectrograph VNIR (400–1000 nm). Calibration set of spectra was extracted from E-20 samples, containing three classes of degradation: class A (optimal quality), class B and class C (maximum deterioration). Reference average spectra were defined for each class. Three models, computed on the calibration set, with a decreasing degree of complexity were compared, according to their ability for segregating leaves at different quality stages (fresh, with incipient and non-visible symptoms of degradation, and degraded): spectral angle mapper distance (SAM), partial least squares discriminant analysis models (PLS-DA), and a non linear index (Leafy Vegetable Evolution, LEVE) combining five wavelengths were included among the previously selected by CovSel procedure. In sets E-10 and E-20, artificial images of the membership degree according to the distance of each pixel to the reference classes, were computed assigning each pixel to the closest reference class. The three methods were able to show the degradation of the leaves with storage time.

Caracterización automatizada de texturas de menas mediante análisis digital de imagen para su aplicación geometalúrgica

La mineralogía de procesos se ha convertido en los últimos años en una herramienta indispensable dentro del ámbito minero-metalúrgico debido fundamentalmente a la emergencia de la Geometalurgia. Esta disciplina en auge, a través de la integración de datos geológicos, mineros y metalúrgicos, proporciona la información necesaria para que el circuito de concentración mineral pueda responder de manera rápida y eficaz a la variabilidad mineralógica inherente a la geología del yacimiento. Para la generación del modelo geometalúrgico, la mineralogía de procesos debe aportar datos cuantitativos sobre los rasgos mineralógicos influyentes en el comportamiento de los minerales y para ello se apoya en el uso de sistemas de análisis mineralógico automatizado. Estos sistemas son capaces de proporcionar gran cantidad de datos mineralógicos de manera rápida y precisa. Sin embargo, cuando se trata de la caracterización de la textura, el mineralogista debe recurrir a descripciones cualitativas basadas en la observación, ya que los sistemas actuales no ofrecen información textural automatizada. Esta tesis doctoral surge precisamente para proporcionar de manera sistemática información textural relevante para los procesos de concentración mineral. La tesis tiene como objetivo principal la identificación y caracterización del tipo de intercrecimiento que un determinado mineral presenta en las partículas minerales, e inicialmente se han tenido en cuenta los siete tipos de intercrecimiento considerados como los más relevantes bajo el punto de vista del comportamiento de las partículas minerales durante flotación, lixiviación y molienda. Para alcanzar este objetivo se ha desarrollado una metodología basada en el diseño y cálculo de una serie de índices numéricos, a los que se ha llamado índices mineralúrgicos, que cumplen una doble función: por un lado, cada índice aporta información relevante para caracterizar los principales rasgos mineralógicos que gobiernan el comportamiento de las partículas minerales a lo largo de los procesos de concentración y por otro lado, estos índices sirven como variables discriminantes para identificar el tipo de intercrecimiento mineral mediante la aplicación de Análisis Discriminante. Dentro del conjunto de índices propuestos en este trabajo, se han considerado algunos índices propuestos por otros autores para su aplicación tanto en el ámbito de la mineralogía como en otros ámbitos de la ciencia de materiales. Se trata del Índice de Contigüidad (Gurland, 1958), Índice de Intercrecimiento (Amstutz y Giger, 1972) e Índice de Coordinación (Jeulin, 1981), adaptados en este caso para el análisis de partículas minerales. El diseño de los índices se ha basado en los principios básicos de la Estereología y el análisis digital de imagen, y su cálculo se ha llevado a cabo aplicando el método de interceptos lineales mediante la programación en MATLAB de varias rutinas. Este método estereológico permite recoger una serie de medidas a partir de las que es posible calcular varios parámetros, tanto estereológicos como geométricos, que han servido de base para calcular los índices mineralúrgicos. Para evaluar la capacidad discriminatoria de los índices mineralúrgicos se han seleccionado 200 casos en los que se puede reconocer de manera clara alguno de los siete tipos de intercrecimiento considerados inicialmente en este trabajo. Para cada uno de estos casos se han calculado los índices mineralúrgicos y se ha aplicado Análisis Discriminante, obteniendo un porcentaje de acierto en la clasificación del 95%. Esta cifra indica que los índices propuestos son discriminadores fiables del tipo de intercrecimiento. Una vez probada la capacidad discriminatoria de los índices, la metodología desarrollada ha sido aplicada para caracterizar una muestra de un concentrado de cobre procedente de la mina Kansanshi (Zambia). Esta caracterización se ha llevado a cabo para obtener la distribución de calcopirita según su tipo de intercrecimiento. La utilidad de esta distribución ha sido analizada bajo diferentes puntos de vista y en todos ellos los índices mineralúrgicos aportan información valiosa para caracterizar el comportamiento mineralúrgico de las partículas minerales. Los resultados derivados tanto del Análisis Discriminante como de la caracterización del concentrado de Kansanshi muestran la fiabilidad, utilidad y versatilidad de la metodología desarrollada, por lo que su integración como herramienta rutinaria en los sistemas actuales de análisis mineralógico pondría a disposición del mineralurgista gran cantidad de información textural complementaria a la información ofrecida por las técnicas actuales de caracterización mineralógica. ABSTRACT Process mineralogy has become in the last decades an essential tool in the mining and metallurgical sphere, especially driven by the emergence of Geometallurgy. This emergent discipline provides required information to efficiently tailor the circuit performance to the mineralogical variability inherent to ore deposits. To contribute to the Geometallurgical model, process mineralogy must provide quantitative data about the main mineralogical features implied in the minerallurgical behaviour of minerals. To address this characterisation, process mineralogy relies on automated systems. These systems are capable of providing a large amount of data quickly and accurately. However, when it comes to the characterisation of texture, mineralogists need to turn to qualitative descriptions based on observation, due to the fact that current systems can not offer quantitative textural information in a routine way. Aiming at the automated characterisation of textural information, this doctoral thesis arises to provide textural information relevant for concentration processes in a systematic way. The main objective of the thesis is the automated identification and characterisation of intergrowth types in mineral particles. Initially, the seven intergrowth types most relevant for flotation, leaching and grinding are considered. To achieve this goal, a methodology has been developed based on the computation of a set of numerical indices, which have been called minerallurgical indices. These indices have been designed with two main purposes: on the one hand, each index provides information to characterise the main mineralogical features which determine particle behaviour during concentration processes and, on the other hand, these indices are used as discriminant variables for identifying the intergrowth type by Discriminant Analysis. Along with the indices developed in this work, three indices proposed by other authors belonging to different fields of materials science have been also considered after being adapted to the analysis of mineral particles. These indices are Contiguity Index (Gurland, 1958), Intergrowth Index (Amstutz and Giger, 1972) and Coordination Index (Jeulin, 1981). The design of minerallurgical indices is based on the fundamental principles of Stereology and Digital Image Analysis. Their computation has been carried out using the linear intercepts method, implemented by means of MATLAB programming. This stereological method provides a set of measurements to obtain several parameters, both stereological and geometric. Based on these parameters, minerallurgical indices have been computed. For the assessment of the discriminant capacity of the developed indices, 200 cases have been selected according to their internal structure, so that one of the seven intergrowth types initially considered in this work can be easily recognised in any of their constituents. Minerallurgical indices have been computed for each case and used as discriminant variables. After applying discriminant analysis, 95% of the cases were correctly classified. This result shows that the proposed indices are reliable identifiers of intergrowth type. Once the discriminant power of the indices has been assessed, the developed methodology has been applied to characterise a copper concentrate sample from the Kansanshi copper mine (Zambia). This characterisation has been carried out to quantify the distribution of chalcopyrite with respect to intergrowth types. Different examples of the application of this distribution have been given to test the usefulness of the method. In all of them, the proposed indices provide valuable information to characterise the minerallurgical behaviour of mineral particles. Results derived from both Discriminant Analysis and the characterisation of the Kansanshi concentrate show the reliability, usefulness and versatility of the developed methodology. Therefore, its integration as a routine tool in current systems of automated mineralogical analysis should make available for minerallurgists a great deal of complementary information to treat the ore more efficiently.

Optical detection of mealiness in apples using laser TDRS.

Mealiness is a textural attribute related to an internal fruit disorder that involves quality loss. It is characterised by the combination of abnormal softness of the fruit and absence of free juiciness in the mouth when eaten by the consumer. Recent research concluded with the development of precise instrumental procedure to measure a scale of mealiness based on the combination of several rheological properties and empirical magnitudes. In this line, time-domain laser reflectance spectroscopy (TDRS) is a new medical technology, used to characterise the optical properties of tissues, and to locate affected areas like tumours. Among its advantages compared to more traditional spectroscopic techniques, there is the feasibility to asses simultaneously and independently two optical parameters: the absorption of the light inside the irradiated body, and the scattering of the photons across the tissues, at each wavelength, generating two coefficients (µa, absorption coeff.; and µ's, transport scattering coeff.). If it is assumed that they are related respectively to chemical components and to physical properties of the sample, TDRS can be applied to the quantification of chemicals and the measurement of the rheological properties (i.e. mealiness estimation) at the same time. Using VIS & NIR lasers as light sources, TDRS was applied in this work to Golden Delicious and Cox apples (n=90), conforming several batches of untreated samples and storage-treated (20°C & 95%RH) to promote the development of mealiness. The collected database was clustered into different groups according to their instrumental test values (Barreiro et al, 1998). The optical coefficients were used as explanatory variables when building discriminant analysis functions for mealiness, achieving a classification score above 80% of correctly identified mealy versus fresh apples.

Automated characterisation of intergrowth textures in mineral particles. A case study

The characterisation of mineral texture has been a major concern for process mineralogists, as liberation characteristics of the ores are intimately related to the mineralogical texture. While a great effort has been done to automatically characterise texture in unbroken ores, the characterisation of textural attributes in mineral particles is usually descriptive. However, the quantitative characterisation of texture in mineral particles is essential to improve and predict the performance of minerallurgical processes (i.e. all the processes involved in the liberation and separation of the mineral of interest) and to achieve a more accurate geometallurgical model. Driven by this necessity of achieving a more complete characterisation of textural attributes in mineral particles, a methodology has been recently developed to automatically characterise the type of intergrowth between mineral phases within particles by means of digital image analysis. In this methodology, a set ofminerallurgical indices has been developed to quantify different mineralogical features and to identify the intergrowth pattern by discriminant analysis. The paper shows the application of the methodology to the textural characterisation of chalcopyrite in the rougher concentrate of the Kansanshi copper mine (Zambia). In this sample, the variety of intergrowth patterns of chalcopyrite with the other minerals has been used to illustrate the methodology. The results obtained show that the method identifies the intergrowth type and provides quantitative information to achieve a complete and detailed mineralogical characterisation. Therefore, the use of this methodology as a routinely tool in automated mineralogy would contribute to a better understanding of the ore behaviour during liberation and separation processes.

Multiple proportion case-basing driven CBRE and its application in the evaluation of possible failure of firms

Case-based reasoning (CBR) is a unique tool for the evaluation of possible failure of firms (EOPFOF) for its eases of interpretation and implementation. Ensemble computing, a variation of group decision in society, provides a potential means of improving predictive performance of CBR-based EOPFOF. This research aims to integrate bagging and proportion case-basing with CBR to generate a method of proportion bagging CBR for EOPFOF. Diverse multiple case bases are first produced by multiple case-basing, in which a volume parameter is introduced to control the size of each case base. Then, the classic case retrieval algorithm is implemented to generate diverse member CBR predictors. Majority voting, the most frequently used mechanism in ensemble computing, is finally used to aggregate outputs of member CBR predictors in order to produce final prediction of the CBR ensemble. In an empirical experiment, we statistically validated the results of the CBR ensemble from multiple case bases by comparing them with those of multivariate discriminant analysis, logistic regression, classic CBR, the best member CBR predictor and bagging CBR ensemble. The results from Chinese EOPFOF prior to 3 years indicate that the new CBR ensemble, which significantly improved CBRs predictive ability, outperformed all the comparative methods.

Mulch materials in processing tomato: a multivariate approach.

Mulch materials of different origins have been introduced into the agricultural sector in recent years alternatively to the standard polyethylene due to its environmental impact. This study aimed to evaluate the multivariate response of mulch materials over three consecutive years in a processing tomato (Solanum lycopersicon L.) crop in Central Spain. Two biodegradable plastic mulches (BD1, BD2), one oxo-biodegradable material (OB), two types of paper (PP1, PP2), and one barley straw cover (BS) were compared using two control treatments (standard black polyethylene [PE] and manual weed control [MW]). A total of 17 variables relating to yield, fruit quality, and weed control were investigated. Several multivariate statistical techniques were applied, including principal component analysis, cluster analysis, and discriminant analysis. A group of mulch materials comprised of OB and BD2 was found to be comparable to black polyethylene regarding all the variables considered. The weed control variables were found to be an important source of discrimination. The two paper mulches tested did not share the same treatment group membership in any case: PP2 presented a multivariate response more similar to the biodegradable plastics, while PP1 was more similar to BS and MW. Based on our multivariate approach, the materials OB and BD2 can be used as an effective, more environmentally friendly alternative to polyethylene mulches.

Age regression from soft aligned face images using low computational resources

The initial step in most facial age estimation systems consists of accurately aligning a model to the output of a face detector (e.g. an Active Appearance Model). This fitting process is very expensive in terms of computational resources and prone to get stuck in local minima. This makes it impractical for analysing faces in resource limited computing devices. In this paper we build a face age regressor that is able to work directly on faces cropped using a state-of-the-art face detector. Our procedure uses K nearest neighbours (K-NN) regression with a metric based on a properly tuned Fisher Linear Discriminant Analysis (LDA) projection matrix. On FG-NET we achieve a state-of-the-art Mean Absolute Error (MAE) of 5.72 years with manually aligned faces. Using face images cropped by a face detector we get a MAE of 6.87 years in the same database. Moreover, most of the algorithms presented in the literature have been evaluated on single database experiments and therefore, they report optimistically biased results. In our cross-database experiments we get a MAE of roughly 12 years, which would be the expected performance in a real world application.

Estimación eficiente de atributos demográficos del rostro humano en imágenes

Sin duda, el rostro humano ofrece mucha más información de la que pensamos. La cara transmite sin nuestro consentimiento señales no verbales, a partir de las interacciones faciales, que dejan al descubierto nuestro estado afectivo, actividad cognitiva, personalidad y enfermedades. Estudios recientes [OFT14, TODMS15] demuestran que muchas de nuestras decisiones sociales e interpersonales derivan de un previo análisis facial de la cara que nos permite establecer si esa persona es confiable, trabajadora, inteligente, etc. Esta interpretación, propensa a errores, deriva de la capacidad innata de los seres humanas de encontrar estas señales e interpretarlas. Esta capacidad es motivo de estudio, con un especial interés en desarrollar métodos que tengan la habilidad de calcular de manera automática estas señales o atributos asociados a la cara. Así, el interés por la estimación de atributos faciales ha crecido rápidamente en los últimos años por las diversas aplicaciones en que estos métodos pueden ser utilizados: marketing dirigido, sistemas de seguridad, interacción hombre-máquina, etc. Sin embargo, éstos están lejos de ser perfectos y robustos en cualquier dominio de problemas. La principal dificultad encontrada es causada por la alta variabilidad intra-clase debida a los cambios en la condición de la imagen: cambios de iluminación, oclusiones, expresiones faciales, edad, género, etnia, etc.; encontradas frecuentemente en imágenes adquiridas en entornos no controlados. Este de trabajo de investigación estudia técnicas de análisis de imágenes para estimar atributos faciales como el género, la edad y la postura, empleando métodos lineales y explotando las dependencias estadísticas entre estos atributos. Adicionalmente, nuestra propuesta se centrará en la construcción de estimadores que tengan una fuerte relación entre rendimiento y coste computacional. Con respecto a éste último punto, estudiamos un conjunto de estrategias para la clasificación de género y las comparamos con una propuesta basada en un clasificador Bayesiano y una adecuada extracción de características. Analizamos en profundidad el motivo de porqué las técnicas lineales no han logrado resultados competitivos hasta la fecha y mostramos cómo obtener rendimientos similares a las mejores técnicas no-lineales. Se propone un segundo algoritmo para la estimación de edad, basado en un regresor K-NN y una adecuada selección de características tal como se propuso para la clasificación de género. A partir de los experimentos desarrollados, observamos que el rendimiento de los clasificadores se reduce significativamente si los ´estos han sido entrenados y probados sobre diferentes bases de datos. Hemos encontrado que una de las causas es la existencia de dependencias entre atributos faciales que no han sido consideradas en la construcción de los clasificadores. Nuestro resultados demuestran que la variabilidad intra-clase puede ser reducida cuando se consideran las dependencias estadísticas entre los atributos faciales de el género, la edad y la pose; mejorando el rendimiento de nuestros clasificadores de atributos faciales con un coste computacional pequeño. Abstract Surely the human face provides much more information than we think. The face provides without our consent nonverbal cues from facial interactions that reveal our emotional state, cognitive activity, personality and disease. Recent studies [OFT14, TODMS15] show that many of our social and interpersonal decisions derive from a previous facial analysis that allows us to establish whether that person is trustworthy, hardworking, intelligent, etc. This error-prone interpretation derives from the innate ability of human beings to find and interpret these signals. This capability is being studied, with a special interest in developing methods that have the ability to automatically calculate these signs or attributes associated with the face. Thus, the interest in the estimation of facial attributes has grown rapidly in recent years by the various applications in which these methods can be used: targeted marketing, security systems, human-computer interaction, etc. However, these are far from being perfect and robust in any domain of problems. The main difficulty encountered is caused by the high intra-class variability due to changes in the condition of the image: lighting changes, occlusions, facial expressions, age, gender, ethnicity, etc.; often found in images acquired in uncontrolled environments. This research work studies image analysis techniques to estimate facial attributes such as gender, age and pose, using linear methods, and exploiting the statistical dependencies between these attributes. In addition, our proposal will focus on the construction of classifiers that have a good balance between performance and computational cost. We studied a set of strategies for gender classification and we compare them with a proposal based on a Bayesian classifier and a suitable feature extraction based on Linear Discriminant Analysis. We study in depth why linear techniques have failed to provide competitive results to date and show how to obtain similar performances to the best non-linear techniques. A second algorithm is proposed for estimating age, which is based on a K-NN regressor and proper selection of features such as those proposed for the classification of gender. From our experiments we note that performance estimates are significantly reduced if they have been trained and tested on different databases. We have found that one of the causes is the existence of dependencies between facial features that have not been considered in the construction of classifiers. Our results demonstrate that intra-class variability can be reduced when considering the statistical dependencies between facial attributes gender, age and pose, thus improving the performance of our classifiers with a reduced computational cost.

Detection of Biological CO2 and 1,3-Pentadiene Using Non-refrigerated Low-Cost MWIR Detectors

The early detection of spoiling metabolic products in contaminated food is a very important tool to control quality. Some volatile compounds produce unpleasant odours at very low concentrations, making their early detection very challenging. This is the case of 1,3-pentadiene produced by microorganisms through decarboxylation of the preservative sorbate. In this work, we have developed a methodology to use the data produced by a low-cost, compact MWIR (Mid-Wave IR) spectrometry device without moving parts, which is based on a linear array of 128 elements of VPD PbSe coupled to a linear variable filter (LVF) working in the spectral range between 3 and 4.6 ?m. This device is able to analyze food headspace gases through dedicated sample presentation setup. This methodology enables the detection of CO2 and the volatile compound 1,3-pentadiene, as compared to synthetic patrons. Data analysis is based on an automated multidimensional dynamic processing of the MWIR spectra. Principal component and discriminant analysis allow segregating between four yeast strains including producers and no producers. The segregation power is accounted as a measure of the discrimination quality.

Nonparametric Message Passing Methods for Cooperative Localization and Tracking

The objective of this thesis is the development of cooperative localization and tracking algorithms using nonparametric message passing techniques. In contrast to the most well-known techniques, the goal is to estimate the posterior probability density function (PDF) of the position of each sensor. This problem can be solved using Bayesian approach, but it is intractable in general case. Nevertheless, the particle-based approximation (via nonparametric representation), and an appropriate factorization of the joint PDFs (using message passing methods), make Bayesian approach acceptable for inference in sensor networks. The well-known method for this problem, nonparametric belief propagation (NBP), can lead to inaccurate beliefs and possible non-convergence in loopy networks. Therefore, we propose four novel algorithms which alleviate these problems: nonparametric generalized belief propagation (NGBP) based on junction tree (NGBP-JT), NGBP based on pseudo-junction tree (NGBP-PJT), NBP based on spanning trees (NBP-ST), and uniformly-reweighted NBP (URW-NBP). We also extend NBP for cooperative localization in mobile networks. In contrast to the previous methods, we use an optional smoothing, provide a novel communication protocol, and increase the efficiency of the sampling techniques. Moreover, we propose novel algorithms for distributed tracking, in which the goal is to track the passive object which cannot locate itself. In particular, we develop distributed particle filtering (DPF) based on three asynchronous belief consensus (BC) algorithms: standard belief consensus (SBC), broadcast gossip (BG), and belief propagation (BP). Finally, the last part of this thesis includes the experimental analysis of some of the proposed algorithms, in which we found that the results based on real measurements are very similar with the results based on theoretical models.

Permutation tests for nonparametric detection

In this paper, the authors provide a methodology to design nonparametric permutation tests and, in particular, nonparametric rank tests for applications in detection. In the first part of the paper, the authors develop the optimization theory of both permutation and rank tests in the Neyman?Pearson sense; in the second part of the paper, they carry out a comparative performance analysis of the permutation and rank tests (detectors) against the parametric ones in radar applications. First, a brief review of some contributions on nonparametric tests is realized. Then, the optimum permutation and rank tests are derived. Finally, a performance analysis is realized by Monte-Carlo simulations for the corresponding detectors, and the results are shown in curves of detection probability versus signal-to-noise ratio

Optimizing panchromatic image change detection based on change index multiband image analysis

This work proposes an optimization of a semi-supervised Change Detection methodology based on a combination of Change Indices (CI) derived from an image multitemporal data set. For this purpose, SPOT 5 Panchromatic images with 2.5 m spatial resolution have been used, from which three Change Indices have been calculated. Two of them are usually known indices; however the third one has been derived considering the Kullbak-Leibler divergence. Then, these three indices have been combined forming a multiband image that has been used in as input for a Support Vector Machine (SVM) classifier where four different discriminant functions have been tested in order to differentiate between change and no_change categories. The performance of the suggested procedure has been assessed applying different quality measures, reaching in each case highly satisfactory values. These results have demonstrated that the simultaneous combination of basic change indices with others more sophisticated like the Kullback-Leibler distance, and the application of non-parametric discriminant functions like those employees in the SVM method, allows solving efficiently a change detection problem.

Region-based moving object detection using spatially conditioned nonparametric models in a GPU

A novel GPU-based nonparametric moving object detection strategy for computer vision tools requiring real-time processing is proposed. An alternative and efficient Bayesian classifier to combine nonparametric background and foreground models allows increasing correct detections while avoiding false detections. Additionally, an efficient region of interest analysis significantly reduces the computational cost of the detections.