Automatic image-based identification of Saimaa ringed seals

The Saimaa ringed seal is one of the most endangered seals in the world. It is a symbol of Lake Saimaa and a lot of effort have been applied to save it. Traditional methods of seal monitoring include capturing the animals and installing sensors on their bodies. These invasive methods for identifying can be painful and affect the behavior of the animals. Automatic identification of seals using computer vision provides a more humane method for the monitoring. This Master's thesis focuses on automatic image-based identification of the Saimaa ringed seals. This consists of detection and segmentation of a seal in an image, analysis of its ring patterns, and identification of the detected seal based on the features of the ring patterns. The proposed algorithm is evaluated with a dataset of 131 individual seals. Based on the experiments with 363 images, 81\% of the images were successfully segmented automatically. Furthermore, a new approach for interactive identification of Saimaa ringed seals is proposed. The results of this research are a starting point for future research in the topic of seal photo-identification.

Reactions to peripheral image motion using a head/eye platform

The authors demonstrate four real-time reactive responses to movement in everyday scenes using an active head/eye platform. They first describe the design and realization of a high-bandwidth four-degree-of-freedom head/eye platform and visual feedback loop for the exploration of motion processing within active vision. The vision system divides processing into two scales and two broad functions. At a coarse, quasi-peripheral scale, detection and segmentation of new motion occurs across the whole image, and at fine scale, tracking of already detected motion takes place within a foveal region. Several simple coarse scale motion sensors which run concurrently at 25 Hz with latencies around 100 ms are detailed. The use of these sensors are discussed to drive the following real-time responses: (1) head/eye saccades to moving regions of interest; (2) a panic response to looming motion; (3) an opto-kinetic response to continuous motion across the image and (4) smooth pursuit of a moving target using motion alone.

Land cover classification of Lago Grande de Curuai floodplain (Amazon, Brazil) using multi-sensor and image fusion techniques

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Development of oil spill detection techniques for satellite optical sensors and their application to monitor oil spill discharge in Mediterranean sea

Satellite remote sensing has proved to be an effective support in timely detection and monitoring of marine oil pollution, mainly due to illegal ship discharges. In this context, we have developed a new methodology and technique for optical oil spill detection, which make use of MODIS L2 and MERIS L1B satellite top of atmosphere (TOA) reflectance imagery, for the first time in a highly automated way. The main idea was combining wide swaths and short revisit times of optical sensors with SAR observations, generally used in oil spill monitoring. This arises from the necessity to overcome the SAR reduced coverage and long revisit time of the monitoring area. This can be done now, given the MODIS and MERIS higher spatial resolution with respect to older sensors (250-300 m vs. 1 km), which consents the identification of smaller spills deriving from illicit discharge at sea. The procedure to obtain identifiable spills in optical reflectance images involves removal of oceanic and atmospheric natural variability, in order to enhance oil-water contrast; image clustering, which purpose is to segment the oil spill eventually presents in the image; finally, the application of a set of criteria for the elimination of those features which look like spills (look-alikes). The final result is a classification of oil spill candidate regions by means of a score based on the above criteria.

Plasmons as sensors

“Plasmon” is a synonym for collective oscillations of the conduction electrons in a metal nanoparticle (excited by an incoming light wave), which cause strong optical responses like efficient light scattering. The scattering cross-section with respect to the light wavelength depends not only on material, size and shape of the nanoparticle, but also on the refractive index of the embedding medium. For this reason, plasmonic nanoparticles are interesting candidates for sensing applications. Here, two novel setups for rapid spectral investigations of single nanoparticles and different sensing experiments are presented.rnrnPrecisely, the novel setups are based on an optical microscope operated in darkfield modus. For the fast single particle spectroscopy (fastSPS) setup, the entrance pinhole of a coupled spectrometer is replaced by a liquid crystal device (LCD) acting as spatially addressable electronic shutter. This improvement allows the automatic and continuous investigation of several particles in parallel for the first time. The second novel setup (RotPOL) usesrna rotating wedge-shaped polarizer and encodes the full polarization information of each particle within one image, which reveals the symmetry of the particles and their plasmon modes. Both setups are used to observe nanoparticle growth in situ on a single-particle level to extract quantitative data on nanoparticle growth.rnrnUsing the fastSPS setup, I investigate the membrane coating of gold nanorods in aqueous solution and show unequivocally the subsequent detection of protein binding to the membrane. This binding process leads to a spectral shift of the particles resonance due to the higher refractive index of the protein compared to water. Hence, the nanosized addressable sensor platform allows for local analysis of protein interactions with biological membranes as a function of the lateral composition of phase separated membranes.rnrnThe sensitivity on changes in the environmental refractive index depends on the particles’ aspect ratio. On the basis of simulations and experiments, I could present the existence of an optimal aspect ratio range between 3 and 4 for gold nanorods for sensing applications. A further sensitivity increase can only be reached by chemical modifications of the gold nanorods. This can be achieved by synthesizing an additional porous gold cage around the nanorods, resulting in a plasmon sensitivity raise of up to 50 % for those “nanorattles” compared to gold nanorods with the same resonance wavelength. Another possibility isrnto coat the gold nanorods with a thin silver shell. This reduces the single particle’s resonance spectral linewidth about 30 %, which enlarges the resolution of the observable shift. rnrnThis silver coating evokes the interesting effect of reducing the ensemble plasmon linewidth by changing the relation connecting particle shape and plasmon resonance wavelength. This change, I term plasmonic focusing, leads to less variation of resonance wavelengths for the same particle size distribution, which I show experimentally and theoretically.rnrnIn a system of two coupled nanoparticles, the plasmon modes of the transversal and longitudinal axis depend on the refractive index of the environmental solution, but only the latter one is influenced by the interparticle distance. I show that monitoring both modes provides a self-calibrating system, where interparticle distance variations and changes of the environmental refractive index can be determined with high precision.

A Multiresolution Image Alignment Technique Based on Direct Methods for Pose Estimation of Aerial Vehicles

In this paper, we seek to expand the use of direct methods in real-time applications by proposing a vision-based strategy for pose estimation of aerial vehicles. The vast majority of approaches make use of features to estimate motion. Conversely, the strategy we propose is based on a MR (Multi-Resolution) implementation of an image registration technique (Inverse Compositional Image Alignment ICIA) using direct methods. An on-board camera in a downwards-looking configuration, and the assumption of planar scenes, are the bases of the algorithm. The motion between frames (rotation and translation) is recovered by decomposing the frame-to-frame homography obtained by the ICIA algorithm applied to a patch that covers around the 80% of the image. When the visual estimation is required (e.g. GPS drop-out), this motion is integrated with the previous known estimation of the vehicles' state, obtained from the on-board sensors (GPS/IMU), and the subsequent estimations are based only on the vision-based motion estimations. The proposed strategy is tested with real flight data in representative stages of a flight: cruise, landing, and take-off, being two of those stages considered critical: take-off and landing. The performance of the pose estimation strategy is analyzed by comparing it with the GPS/IMU estimations. Results show correlation between the visual estimation obtained with the MR-ICIA and the GPS/IMU data, that demonstrate that the visual estimation can be used to provide a good approximation of the vehicle's state when it is required (e.g. GPS drop-outs). In terms of performance, the proposed strategy is able to maintain an estimation of the vehicle's state for more than one minute, at real-time frame rates based, only on visual information.

An image segmentation based on a genetic algorithm for determining soil coverage by crop residues

Determination of the soil coverage by crop residues after ploughing is a fundamental element of Conservation Agriculture. This paper presents the application of genetic algorithms employed during the fine tuning of the segmentation process of a digital image with the aim of automatically quantifying the residue coverage. In other words, the objective is to achieve a segmentation that would permit the discrimination of the texture of the residue so that the output of the segmentation process is a binary image in which residue zones are isolated from the rest. The RGB images used come from a sample of images in which sections of terrain were photographed with a conventional camera positioned in zenith orientation atop a tripod. The images were taken outdoors under uncontrolled lighting conditions. Up to 92% similarity was achieved between the images obtained by the segmentation process proposed in this paper and the templates made by an elaborate manual tracing process. In addition to the proposed segmentation procedure and the fine tuning procedure that was developed, a global quantification of the soil coverage by residues for the sampled area was achieved that differed by only 0.85% from the quantification obtained using template images. Moreover, the proposed method does not depend on the type of residue present in the image. The study was conducted at the experimental farm “El Encín” in Alcalá de Henares (Madrid, Spain).

Comparison of fractal dimensions based on segmented NDVI fields obtained from different remote sensors.

Satellite image data have become an important source of information for monitoring vegetation and mapping land cover at several scales. Beside this, the distribution and phenology of vegetation is largely associated with climate, terrain characteristics and human activity. Various vegetation indices have been developed for qualitative and quantitative assessment of vegetation using remote spectral measurements. In particular, sensors with spectral bands in the red (RED) and near-infrared (NIR) lend themselves well to vegetation monitoring and based on them [(NIR - RED) / (NIR + RED)] Normalized Difference Vegetation Index (NDVI) has been widespread used. Given that the characteristics of spectral bands in RED and NIR vary distinctly from sensor to sensor, NDVI values based on data from different instruments will not be directly comparable. The spatial resolution also varies significantly between sensors, as well as within a given scene in the case of wide-angle and oblique sensors. As a result, NDVI values will vary according to combinations of the heterogeneity and scale of terrestrial surfaces and pixel footprint sizes. Therefore, the question arises as to the impact of differences in spectral and spatial resolutions on vegetation indices like the NDVI. The aim of this study is to establish a comparison between two different sensors in their NDVI values at different spatial resolutions.

Comparison of NDVI fields obtained from different remote sensors

Satellite image data have become an important source of information for monitoring vegetation and mapping land cover at several scales. Beside this, the distribution and phenology of vegetation is largely associated with climate, terrain characteristics and human activity. Various vegetation indices have been developed for qualitative and quantitative assessment of vegetation using remote spectral measurements. In particular, sensors with spectral bands in the red (RED) and near-infrared (NIR) lend themselves well to vegetation monitoring and based on them [(NIR - RED) / (NIR + RED)] Normalized Difference Vegetation Index (NDVI) has been widespread used. Given that the characteristics of spectral bands in RED and NIR vary distinctly from sensor to sensor, NDVI values based on data from different instruments will not be directly comparable. The spatial resolution also varies significantly between sensors, as well as within a given scene in the case of wide-angle and oblique sensors. As a result, NDVI values will vary according to combinations of the heterogeneity and scale of terrestrial surfaces and pixel footprint sizes. Therefore, the question arises as to the impact of differences in spectral and spatial resolutions on vegetation indices like the NDVI and their interpretation as a drought index. During 2012 three locations (at Salamanca, Granada and Córdoba) were selected and a periodic pasture monitoring and botanic composition were achieved. Daily precipitation, temperature and monthly soil water content were measurement as well as fresh and dry pasture weight. At the same time, remote sensing images were capture by DEIMOS-1 and MODIS of the chosen places. DEIMOS-1 is based on the concept Microsat-100 from Surrey. It is conceived for obtaining Earth images with a good enough resolution to study the terrestrial vegetation cover (20x20 m), although with a great range of visual field (600 km) in order to obtain those images with high temporal resolution and at a reduced cost. By contranst, MODIS images present a much lower spatial resolution (500x500 m). The aim of this study is to establish a comparison between two different sensors in their NDVI values at different spatial resolutions. Acknowledgements. This work was partially supported by ENESA under project P10 0220C-823. Funding provided by Spanish Ministerio de Ciencia e Innovación (MICINN) through project no. MTM2009-14621 and i-MATH No. CSD2006-00032 is greatly appreciated.

Evidence for reciprocal antagonism between motion sensors tuned to coarse and fine features

Early visual processing analyses fine and coarse image features separately. Here we show that motion signals derived from fine and coarse analyses are combined in rather a surprising way: Coarse and fine motion sensors representing the same direction of motion inhibit one another and an imbalance can reverse the motion perceived. Observers judged the direction of motion of patches of filtered two-dimensional noise, centered on 1 and 3 cycles/deg. When both sets of noise were present and only the 3 cycles/deg noise moved, judgments were reversed at short durations. When both sets of noise moved, judgments were correct but sensitivity was impaired. Reversals and impairments occurred both with isotropic noise and with orientation-filtered noise. The reversals and impairments could be simulated in a model of motion sensing by adding a stage in which the outputs of motion sensors tuned to 1 and 3 cycles/deg and the same direction of motion were subtracted from one another. The subtraction model predicted and we confirmed in experiments with orientation-filtered noise that if the 1 cycle/deg noise flickered and the 3 cycles/deg noise moved, the 1 cycle/deg noise appeared to move in the opposite direction to the 3 cycles/deg noise even at long durations.

Low-cost interrogation of optical fibre Bragg grating sensors

Through the application of novel signal processing techniques we are able to measure physical measurands with both high accuracy and low noise susceptibility. The first interrogation scheme is based upon a CCD spectrometer. We compare different algorithms for resolving the Bragg wavelength from a low resolution discrete representation of the reflected spectrum, and present optimal processing methods for providing a high integrity measurement from the reflection image. Our second sensing scheme uses a novel network of sensors to measure the distributive strain response of a mechanical system. Using neural network processing methods we demonstrate the measurement capabilities of a scalable low-cost fibre Bragg grating sensor network. This network has been shown to be comparable with the performance of existing fibre Bragg grating sensing techniques, at a greatly reduced implementation cost.

Bottleneck Problem Solution using Biological Models of Attention in High Resolution Tracking Sensors

Every high resolution imaging system suffers from the bottleneck problem. This problem relates to the huge amount of data transmission from the sensor array to a digital signal processing (DSP) and to bottleneck in performance, caused by the requirement to process a large amount of information in parallel. The same problem exists in biological vision systems, where the information, sensed by many millions of receptors should be transmitted and processed in real time. Models, describing the bottleneck problem solutions in biological systems fall in the field of visual attention. This paper presents the bottleneck problem existing in imagers used for real time salient target tracking and proposes a simple solution by employing models of attention, found in biological systems. The bottleneck problem in imaging systems is presented, the existing models of visual attention are discussed and the architecture of the proposed imager is shown.

Infrared image enhancement based on hybrid-domain consideration and data fusion methods

Au cours des dernières décennies, l’effort sur les applications de capteurs infrarouges a largement progressé dans le monde. Mais, une certaine difficulté demeure, en ce qui concerne le fait que les objets ne sont pas assez clairs ou ne peuvent pas toujours être distingués facilement dans l’image obtenue pour la scène observée. L’amélioration de l’image infrarouge a joué un rôle important dans le développement de technologies de la vision infrarouge de l’ordinateur, le traitement de l’image et les essais non destructifs, etc. Cette thèse traite de la question des techniques d’amélioration de l’image infrarouge en deux aspects, y compris le traitement d’une seule image infrarouge dans le domaine hybride espacefréquence, et la fusion d’images infrarouges et visibles employant la technique du nonsubsampled Contourlet transformer (NSCT). La fusion d’images peut être considérée comme étant la poursuite de l’exploration du modèle d’amélioration de l’image unique infrarouge, alors qu’il combine les images infrarouges et visibles en une seule image pour représenter et améliorer toutes les informations utiles et les caractéristiques des images sources, car une seule image ne pouvait contenir tous les renseignements pertinents ou disponibles en raison de restrictions découlant de tout capteur unique de l’imagerie. Nous examinons et faisons une enquête concernant le développement de techniques d’amélioration d’images infrarouges, et ensuite nous nous consacrons à l’amélioration de l’image unique infrarouge, et nous proposons un schéma d’amélioration de domaine hybride avec une méthode d’évaluation floue de seuil amélioré, qui permet d’obtenir une qualité d’image supérieure et améliore la perception visuelle humaine. Les techniques de fusion d’images infrarouges et visibles sont établies à l’aide de la mise en oeuvre d’une mise en registre précise des images sources acquises par différents capteurs. L’algorithme SURF-RANSAC est appliqué pour la mise en registre tout au long des travaux de recherche, ce qui conduit à des images mises en registre de façon très précise et des bénéfices accrus pour le traitement de fusion. Pour les questions de fusion d’images infrarouges et visibles, une série d’approches avancées et efficaces sont proposés. Une méthode standard de fusion à base de NSCT multi-canal est présente comme référence pour les approches de fusion proposées suivantes. Une approche conjointe de fusion, impliquant l’Adaptive-Gaussian NSCT et la transformée en ondelettes (Wavelet Transform, WT) est propose, ce qui conduit à des résultats de fusion qui sont meilleurs que ceux obtenus avec les méthodes non-adaptatives générales. Une approche de fusion basée sur le NSCT employant la détection comprime (CS, compressed sensing) et de la variation totale (TV) à des coefficients d’échantillons clairsemés et effectuant la reconstruction de coefficients fusionnés de façon précise est proposée, qui obtient de bien meilleurs résultats de fusion par le biais d’une pré-amélioration de l’image infrarouge et en diminuant les informations redondantes des coefficients de fusion. Une procédure de fusion basée sur le NSCT utilisant une technique de détection rapide de rétrécissement itératif comprimé (fast iterative-shrinking compressed sensing, FISCS) est proposée pour compresser les coefficients décomposés et reconstruire les coefficients fusionnés dans le processus de fusion, qui conduit à de meilleurs résultats plus rapidement et d’une manière efficace.

Tradução, adaptação cultural e validação do questionário Body image After Breast Cancer para a língua portuguesa do Brasil

Universidade Estadual de Campinas . Faculdade de Educação Física