Pasar al contenido principal

AUGMENTED REALITY SUBSEA EXPLORATION ASSISTANT (ARSEA): UNA HERRAMIENTA PARA LA INSPECCION ASISTIDA Y LA RECONSTRUCCION 3D ON-LINE DE ENTORNOS SUBMARINOS

TIN2014-58662-R

•
Nombre agencia financiadora Ministerio de Economía y Competitividad
Acrónimo agencia financiadora MINECO
Programa Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Convocatoria Proyectos I+D+i Retos de Investigación
Año convocatoria 2014
Plan Estatal 2013-2016 Plan Estatal de Investigación Científica y Técnica y de Innovación
Unidad de gestión Secretaría de Estado de Educación, Formación Profesional y Universidades
Centro beneficiario UNIVERSIDAD DE LAS ISLAS BALEARES
Identificador persistente http://dx.doi.org/10.13039/501100003329

Resultados relacionados

Resultados totales (Incluyendo duplicados): 8
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Hacia la Navegación Visual de un Vehículo Autónomo Submarino en Áreas con Posidonia Oceanica, Towards Visual Navigation of an Autonomous Underwater Vehicle in Areas with Posidonia Oceanica

  • Bonin-Font, Francisco
  • Coll Gomila, Carles
  • Oliver Codina, Gabriel
[ES] Este artículo presenta los resultados de un estudio experimental exhaustivo que determina el tipo de características visuales que presentan una mayor robustez, estabilidad y trazabilidad en imágenes submarinas tomadas en entornos colonizados con Posidonia Oceanica (P.O.), sean consecutivas o que cierran bucles (imágenes que muestran una misma área, parcial o totalmente, tomadas en tiempos distintos, desde puntos de vista distintos o incluso en condiciones de iluminación diferentes). El trabajo se ha centrado en dos puntos fundamentales: a) evaluar la capacidad que pueden tener varias técnicas de aumento de contraste en imágenes con P.O. a la hora de aumentar el número y calidad de las características visuales, y b) encontrar la combinación detector/descriptor invariante a rotación y traslación, que maximiza el número de correspondencias inliers usadas posteriormente para el cálculo de la odometria visual, o en el registro de imágenes que cierran bucles., [EN] This paper presents an exhaustive, extensive and detailed experimental assessment of different types of visual key-points in terms of robustness, stability and traceability, in images taken in marine areas densely colonized with Posidonia Oceanica (P.O.). This work has been focused mainly in two issues: a) evaluating the  capacity of several image color and contrast enhancing preprocessing techniques to increase the image quality and the number of stable features, and b) finding the pair feature detector/descriptor, from a wide range of different combinations, that maximizes the number of inlier correspondences in consecutive frames or frames that close a loop (images that overlap, taken at distant time instants, from different viewpoints or even with different environmental conditions). Conclusions extracted from both evaluations will affect directly the quality of visual odometers and/or the image registration processes involved in visual SLAM approaches. , Ministerio de Economía y Competitividad a través del proyecto TIN2014- 58662-R y fondos FEDER
Repositori Institucional UIB
Publicación Artículo científico (JournalArticle)

A Contrario Detection of Faces: A Case Example

  • Lisani, Jose-Luis
  • Ramis, Silvia
  • Perales, Francisco J.
[eng] The a contrario framework is a statistical formulation of a perception principle that permits one to detect meaningful structures in data. It has been applied to the detection of lines and contours in images, moving objects in video, etc., but no attempt has been made to use it for the detection of faces. The goal of this paper is to show that the a contrario formulation can be adapted to the face detection method described by Viola and Jones in their seminal work. We propose an alternative to the cascade of classiers proposed by the authors by introducing a stochastic a contrario model for the detections of a single classier, from which adaptive detection thresholds may be inferred. The result is a single classier whose detection rates are similar to those of a cascade of classiers. Moreover, we show how a very short cascade of classiers can be constructed, which improves the accuracy of a classical cascade, at a much lower computational cost. The results prove the validity of the a contrario approach for face detection and suggest that the same principles might be used to improve the performance of state-of-the-art methods.
Repositori Institucional UIB
Publicación Capítulo o parte de un libro (BookPart)

Towards Multi Session Visual SLAM in Underwater Environments Colonized with Posidonia Oceanica

  • Burguera Burguera, Antonio
  • Bonin-Font, Francisco
[eng] This paper describes a multi-session monocular SLAM approach addressed to underwater environments. It has three main blocks: a) visual odometry, b) loop-closing detection; loop closings inside each individual session are found applying feature matching with RANSAC, and since no geometric relation between images of different sessions is available, multi-session loop closings are detected via an image hash matching procedure, alleviating also the computational cost of the image comparisons, and c) an Iterated Extended Kalman Filter (IEKF) based optimization process used to refine the different individual trajectories and to join the different maps; the global optimization process (map joining) can be delayed until certain number of loop closings are found, reducing the global running time. Several experiments using marine imagery in areas colonized with Posidonia Oceanica show the robustness of this approach.
Repositori Institucional UIB
Publicación Capítulo o parte de un libro (BookPart)

A Novel Approach to Register Sonar Data for Underwater Robot Localization

  • Burguera, Antoni
[eng] This paper proposes a novel approach to perform underwater robot localization using sonar data. The proposal relies on two key points: 1) the processing of the sonar measurements to build probabilistic point clouds; and 2) the registration of these point clouds. These key points are achieved by means of an acoustic model that removes the effects of the uneven ensonification followed by a probabilistic interpretation of the sonar measurements. Also, a registration algorithm is presented and used to match consecutively gathered scans in order to globally localize an autonomous underwater vehicles (AUV).
Repositori Institucional UIB
Publicación Capítulo o parte de un libro (BookPart)

Machine Learning and Deep Learning Strategies to Identify Posidonia Meadows in Underwater Images

  • González-Cid, Yolanda
  • Burguera, Antoni
  • Bonin-Font, Francisco
  • Matamoros, Alejandro
[eng] This paper describes how to automatically identify Posidonea Oceanica (P.O.) from seabed images gathered by a bottom-looking camera. Different methods based on machine learning and deep learning algorithms are presented and compared. On the one hand, texture descriptors and co-occurrence matrices are used to characterize the images and classify the P.O. regions by means of Support Vector Machine and Artificial Neural Networks. On the other hand, Convolutional Neural Networks are used in the Deep Learning approach. The experimental results obtained demonstrate the effectiveness of the algorithms proposed to automatically identify P.O. meadows in underwater images.
Repositori Institucional UIB
Publicación Capítulo o parte de un libro (BookPart)

Building High Resolution Maps of Large Subsea Areas using Side-Scan Sonar

  • Burguera, Antoni
  • Oliver, Gabriel
[eng] The goal of this study is to generate highresolution sea floor maps using a Side-Scan Sonar (SSS) on board of an Autonomous Underwater Vehicle (AUV). This is achieved by explicitly taking into account the SSS operation as follows. First, the raw SSS data is corrected both in intensity, by removing the effects of the uneven ensonification pattern, and range, by projecting the data to the sea floor. Second, the AUV pose is estimated continuously. Third, a probabilistic SSS model is defined and used to estimate the probability of each sea-floor region to be observed. This probabilistic information is then used to weight the contribution of each SSS measurement to the map. Because of these models, arbitrary map resolutions can be achieved, even beyond the sensor resolution.
Repositori Institucional UIB
Publicación Capítulo o parte de un libro (BookPart)

Building Large-Scale Coverage Maps of Posidonia Oceanica using an Autonomous Underwater Vehicle

  • Burguera, Antoni
  • Bonin-Font, Francisco
  • García-Fidalgo, Emilio
[eng] This paper proposes a method to build large scale coverage maps of Posidonia Oceanica (P.O.) using the imagery provided by an AUV endowed with a bottom-looking camera. The process begins by detecting the regions containing P.O. in each individual image. Afterwards, the overlap between images is searched and the P.O. detection data in these overlapping regions is fused to build a global detection map. The paper proposes three different descriptors to detect P.O. in individual images and five strategies to fuse the overlapping regions. The experiments, performed in coastal areas of Mallorca, evaluate and compare the proposed descriptors and fusion techniques, showing that our approach surpasses the 91% detection rate.
Repositori Institucional UIB
Publicación Capítulo o parte de un libro (BookPart)

Towards Automatic Visual Sea Grass Detection in Underwater Areas of Ecological Interest

  • Burguera, Antoni
  • Bonin-Font, Francisco
  • Lisani, José Luis
  • Petro, Ana Belén
  • Oliver, Gabriel
[eng] In areas of ecological interest, the detection and control of seaweed such as Posidonia Oceanica is usually performed by divers. Due to the limited capacity of the scuba tanks and the human security protocols, this task involves several short immersions leading to poor temporal and spatial data resolution. Thus, it is desirable to automate this task by means of underwater robots. This paper describes a method to autonomously detect Posidonia Oceanica in the imagery gathered by an underwater robot. The proposed approach uses a set of Gabor filters to characterize an image. This characterization is used to detect the regions containing seaweed by means of a Support Vector Machine. The experiments, conducted with an Autonomous Underwater Robot in several marine areas of Mallorca, show promising results towards the automated seafloor classification from extended video sequences.