Simultaneous, vision-based fish instance segmentation, species classification and size regression.

Overexploitation of fisheries is a worldwide problem, which is leading to a large loss of diversity, and affects human communities indirectly through the loss of traditional jobs, cultural heritage, etc. To address this issue, governments have started accumulating data on fishing activities, to determine biomass extraction rates, and fisheries status. However, these data are often estimated from small samplings, which can lead to partially inaccurate assessments. Fishing can also benefit of the digitization process that many industries are undergoing. Wholesale fish markets, where vessels disembark, can be the point of contact to retrieve valuable information on biomass extraction rates, and can do so automatically. Fine-grained knowledge about the fish species, quantities, sizes, etc. that are caught can be therefore very valuable to all stakeholders, and particularly decision-makers regarding fisheries conservation, sustainable, and long-term exploitation. In this regard, this article presents a full workflow for fish instance segmentation, species classification, and size estimation from uncalibrated images of fish trays at the fish market, in order to automate information extraction that can be helpful in such scenarios. Our results on fish instance segmentation and species classification show an overall mean average precision (mAP) at 50% intersection-over-union (IoU) of 70.42%, while fish size estimation shows a mean average error (MAE) of only 1.27 cm.

Iterative multilinear optimization for planar model fitting under geometric constraints.

Planes are the core geometric models present everywhere in the three-dimensional real world. There are many examples of manual constructions based on planar patches: facades, corridors, packages, boxes, etc. In these constructions, planar patches must satisfy orthogonal constraints by design (e.g. walls with a ceiling and floor). The hypothesis is that by exploiting orthogonality constraints when possible in the scene, we can perform a reconstruction from a set of points captured by 3D cameras with high accuracy and a low response time. We introduce a method that can iteratively fit a planar model in the presence of noise according to three main steps: a clustering-based unsupervised step that builds pre-clusters from the set of (noisy) points; a linear regression-based supervised step that optimizes a set of planes from the clusters; a reassignment step that challenges the members of the current clusters in a way that minimizes the residuals of the linear predictors. The main contribution is that the method can simultaneously fit different planes in a point cloud providing a good accuracy/speed trade-off even in the presence of noise and outliers, with a smaller processing time compared with previous methods. An extensive experimental study on synthetic data is conducted to compare our method with the most current and representative methods. The quantitative results provide indisputable evidence that our method can generate very accurate models faster than baseline methods. Moreover, two case studies for reconstructing planar-based objects using a Kinect sensor are presented to provide qualitative evidence of the efficiency of our method in real applications.

RGB-D-Based Framework to Acquire, Visualize and Measure the Human Body for Dietetic Treatments.

This research aims to improve dietetic-nutritional treatment using state-of-the-art RGB-D sensors and virtual reality (VR) technology. Recent studies show that adherence to treatment can be improved using multimedia technologies. However, there are few studies using 3D data and VR technologies for this purpose. On the other hand, obtaining 3D measurements of the human body and analyzing them over time (4D) in patients undergoing dietary treatment is a challenging field. The main contribution of the work is to provide a framework to study the effect of 4D body model visualization on adherence to obesity treatment. The system can obtain a complete 3D model of a body using low-cost technology, allowing future straightforward transference with sufficient accuracy and realistic visualization, enabling the analysis of the evolution (4D) of the shape during the treatment of obesity. The 3D body models will be used for studying the effect of visualization on adherence to obesity treatment using 2D and VR devices. Moreover, we will use the acquired 3D models to obtain measurements of the body. An analysis of the accuracy of the proposed methods for obtaining measurements with both synthetic and real objects has been carried out.

Evaluating Impact on Motivation and Academic Performance of a Game-Based Learning Experience Using Kahoot.

Gamification methods adapt the mechanics of games to educational environments for the improvement of the teaching-learning process. Serious games play an important role as tools for gamification, in particular in the context of software engineering courses because of the idiosyncratic nature of the topic. However, the studies on the improvement of student performance resulting from the use of gamification and serious games in courses with different contexts are not conclusive. More empirical research is thus needed to obtain reliable results on the effectiveness, benefits and drawbacks. The overall objective of this work is to study the benefits generated by serious games in the teaching-learning process of Computer Engineering degrees, analyzing the impact on the motivation and student satisfaction, as well as on the learning outcomes and results finally achieved. To this end, an intervention is proposed in the subject of Computer Architecture based on two components covering theoretical and practical sessions. In the theoretical sessions, a serious game experience using Kahoot has been introduced, complementing the master classes and class exercises. For the practical sessions, the development of projects with groups of students has been proposed, whose results in terms of computer performance can be compared through a competition (hackathon). Evaluation of the serious game-based intervention has been approached in terms of student satisfaction and motivation, as well as improved academic performance. In order to assess student satisfaction, surveys have been used to assess the effect on student motivation and satisfaction. For the evaluation of academic performance, a comparative analysis between an experimental and a control group has been carried out, noting a slight increase in the experimental group students' marks.

A Novel Active Imaging Model to Design Visual Systems: A Case of Inspection System for Specular Surfaces.

The use of visual information is a very well known input from different kinds of sensors. However, most of the perception problems are individually modeled and tackled. It is necessary to provide a general imaging model that allows us to parametrize different input systems as well as their problems and possible solutions. In this paper, we present an active vision model considering the imaging system as a whole (including camera, lighting system, object to be perceived) in order to propose solutions to automated visual systems that present problems that we perceive. As a concrete case study, we instantiate the model in a real application and still challenging problem: automated visual inspection. It is one of the most used quality control systems to detect defects on manufactured objects. However, it presents problems for specular products. We model these perception problems taking into account environmental conditions and camera parameters that allow a system to properly perceive the specific object characteristics to determine defects on surfaces. The validation of the model has been carried out using simulations providing an efficient way to perform a large set of tests (different environment conditions and camera parameters) as a previous step of experimentation in real manufacturing environments, which more complex in terms of instrumentation and more expensive. Results prove the success of the model application adjusting scale, viewpoint and lighting conditions to detect structural and color defects on specular surfaces.

A Quantitative Comparison of Calibration Methods for RGB-D Sensors Using Different Technologies.

RGB-D (Red Green Blue and Depth) sensors are devices that can provide color and depth information from a scene at the same time. Recently, they have been widely used in many solutions due to their commercial growth from the entertainment market to many diverse areas (e.g., robotics, CAD, etc.). In the research community, these devices have had good uptake due to their acceptable levelofaccuracyformanyapplicationsandtheirlowcost,butinsomecases,theyworkatthelimitof their sensitivity, near to the minimum feature size that can be perceived. For this reason, calibration processes are critical in order to increase their accuracy and enable them to meet the requirements of such kinds of applications. To the best of our knowledge, there is not a comparative study of calibration algorithms evaluating its results in multiple RGB-D sensors. Specifically, in this paper, a comparison of the three most used calibration methods have been applied to three different RGB-D sensors based on structured light and time-of-flight. The comparison of methods has been carried out by a set of experiments to evaluate the accuracy of depth measurements. Additionally, an object reconstruction application has been used as example of an application for which the sensor works at the limit of its sensitivity. The obtained results of reconstruction have been evaluated through visual inspection and quantitative measurements.

A comparative study of registration methods for RGB-D video of static scenes.

The use of RGB-D sensors for mapping and recognition tasks in robotics or, in general, for virtual reconstruction has increased in recent years. The key aspect of these kinds of sensors is that they provide both depth and color information using the same device. In this paper, we present a comparative analysis of the most important methods used in the literature for the registration of subsequent RGB-D video frames in static scenarios. The analysis begins by explaining the characteristics of the registration problem, dividing it into two representative applications: scene modeling and object reconstruction. Then, a detailed experimentation is carried out to determine the behavior of the different methods depending on the application. For both applications, we used standard datasets and a new one built for object reconstruction.