Ciência da Computação
URI permanente desta comunidadehttps://repositorio.fei.edu.br/handle/FEI/342
Navegar
44 resultados
Resultados da Pesquisa
- Computer Vision Based on a Modular Neural Network for Automatic Assessment of Physical Therapy Rehabilitation Activities(2022-01-05) FRANCISCO, J. A.; Paulo RodriguesAuthorPhysical rehabilitation techniques during the treatment of clinical pathology are one of the most challenging areas for the medical structure, patients, and families. In large and continental countries, remote monitoring of this treatment is essential. However, equipment and medical follow-up during exercises still have high costs. With the improvement of computer vision and machine learning techniques, some computational, less expensive alternatives have been proposed in the literature. However, monitoring patients during physical rehabilitation exercises with the help of artificial intelligence by a health professional, especially from the capture of visual signals, is still a challenge and poorly explored in the scientific-technological literature. This work aims to propose a new methodology based on computer vision and machine learning for remote tracking of the body joints of patients during physiotherapy rehabilitation exercises. As a new contribution, this work presents a modular neural network architecture composed of two modules: one for detecting physical exercises and another for measuring how much is correct. Another contribution is a strategy for expanding databases, considering that generic databases for this type of exercise are rare on the internet. The results showed that both modules obtained more than 90% of accuracy in recognition and their respective validation.
- Discrete-event simulation of an irrigation system using Internet of Things(2022-06-01) GOMES ALVES, R.; Rodrigo Maia; Fabio Lima© 2003-2012 IEEE.Agricultural water consumption represents 69% of all freshwater used on the planet. In addition, it is necessary to increase food production by 50% by 2050. The use of Internet of Things platforms to carry out the sensing and monitoring of the agricultural environment is increasingly present in the literature. One of the difficulties that such platforms face is to validate the platform's operation in different irrigation systems, as it is often necessary for specialists to work in the connection of sensors and actuators that already exist on farms or that are defined in the design of such systems. Within this context, a discrete-event simulation of an irrigation system integrated into an Internet of Things platform was developed in this work. The digital manufacturing software Plant Simulation was used to perform the discrete event simulation. An OPC UA server establishes real-time communication between the Internet of Things platform and the simulation software. Thus, farmers may verify, in real-time, how a given irrigation prescription, sent by the IoT platform, takes place in the irrigation system.
- A bipartite graph approach to retrieve similar 3D models with different resolution and types of cardiomyopathies(2022-05-01) Leila Bergamasco; LIMA, K.R.P.S.; ROCHITTE, C. E.; NUNES, F. L. S.Three-dimensional (3D) model retrieval uses content-based image retrieval (CBIR) techniques to search for the most similar 3D objects in a dataset, usually considering their geometry and organization in a feature vector. Feature vectors from different objects were compared to establish their similarities. Although this type of comparison typically uses metric distances, such metrics present limitations when the vector lengths are different. Signal-based descriptors are a promising approach for extracting features from 3D objects, but they generate feature vectors with different lengths. Thus, new methods for measuring the similarity are required. This study proposes an approach to 3D model retrieval as a network flow problem using bipartite graphs. The approach was applied to support the diagnosis of cardiomyopathies, considering 3D objects reconstructed from cardiac images of the left ventricle. We achieved an AUC value of 0.93 under the best retrieval scenario. The results also indicate that modeling a 3D model retrieval technique as a network flow problem using graphs can provide a promising manner to compare 3D objects with different shapes and sizes. This strategy, coupled with personal patient data, achieves better results than methods using classical comparison approaches.
- A new approach based on computer vision and non-linear Kalman filtering to monitor the nebulization quality of oil flames(2013-09-15) FLEURY, A. T.; TRIGO, F. C.; MARTINS, F. P. R.The nebulization quality of oil flames, an important characteristic exhibited by combustion processes of petroleum refinery furnaces, is mostly affected by variations on the values ofthe vapor flow rate (VFR). Expressive visual changes in the flame patterns and decay of the combustion efficiency are observed when the process is tuned by diminishing the VFR. Such behavior is supported by experimental evidence showing that too low values of VFR and solid particulate material rate increase are strongly correlate d. Given the economical importance of keeping this parameter under control, a laborator ial vertical furnace was devised with the purpose of carrying out experiments to prototype acomputer vision system capable of estimati ng VFR values through the examination of test charact eristic vectors based on geometric properties of the grey level histogram of instantaneous flame images. Firstly, atraining set composed of feature vectors from all the images collected during experiments with a priori known VFR values are properly organized and analgorithm is applied to this data in order to generate a fuzzy measurement vector whose components represent membership degrees to the 'high nebulization quality'fuzzy set. Fuzzy classification vectors from images with unknown a priori VFR values are, then, assumed tobe state-vectors inarandom-walk model, and a non-linear Tikhonov regularized Kalman filter is applied to estimate the state and the corresponding nebulization quality. The successful validation of the output data, even based onsmall training data sets, indicates that the proposed approach could beapplied to synthesize a real-time algorithm for evaluating the nebulization quality of combustion processes in petroleum refinery furnaces that use oil flamesasthe heating source. © 2013 Elsevier Ltd. All rights reserved.
- Human Factors Aspects in Test Cases Formalization(2015-01-05) CASTRO, B. P. P. D.; Plinio Thomaz Aquino Junior© 2015 The AuthorsThis article describes a method capable of automate the process of case tests documentation, using as its implantation format, a flexible approach where companies gradually adopt the method's characteristics and requirements. The research and its experiments follows a usability testing approach, focused in human factors to build the test data. The method's execution occurs through business requirements documentation based on scenarios. These data are classified and organized in order to obtain standard testable objects. These test items are documented in test cases and the process starts again using the test case as basis. The cyclic process repeats until the test conditions have been validated.
- Identification of the state-space dynamics of oil flames through computer vision and modal techniques(2015-04-01) SILVA, R. P.; FLEURY, A. T.; MARTINS F. P. R.; PONGE-FERREIRA, W. J. A.; TRIGO, F. C.© 2014 Elsevier Ltd.In industrial oil furnaces, unstable flames can lead to potentially dangerous conditions. For this reason, elaborate control systems are used to monitor the various parameters of the process that could become the source of such problems. A current trend in research is the one that seeks to apply artificial intelligence techniques to efficiently identify a priory anomalous behavior of the flames, so as to help improving the time response of the automatic control. In system dynamics theory, it is common sense that an accurate modeling of the process under study directly affects the performance of the controlling apparatus. Unfortunately, due to the complexity of the process, physical models of flame propagation are still not as much faithful as they should to be used for control purposes. On the other hand, could the complex dynamics of flame propagation be described in terms of an identified assumed model, one would come up with a tool for the improvement of the control strategy. In this work, a new approach based on Operational Modal Analysis (OMA) tools is used to identify four degree-of-freedom second order state-space models of oil flame dynamics in a prototype furnace. Grabbed images of a CCD camera, after being processed through a computer vision method, provide sets of characteristic vectors which, then, serve as input data to an identification OMA algorithm based on the Ibrahim Time Domain Method. Models of unstable and stable flames are built and validated through spectral analysis of the reconstructed time-domain characteristic vectors. The truthfulness of the validation scheme was then confirmed by a quantitative modal assurance criterion modified to suit the current application. On the grounds of the results obtained, it is possible to assert that the proposed approach for the description of flame dynamics can likely predict the occurrence of unstable conditions, thus becoming another tool that might be used in an automated control system.
- Cognitive brain mapping of auditors and accountants in going concern judgments Mapeamento cognitivo cerebral de auditores e contadores em julgamentos de continuidade operacional(2017-01-05) CARVALHO JÚNIOR, C. V. DE O.; CORNACCHIONE, E.ROCHA, A. F. DE; ROCHA, A. F. DE; ROCHA, F. T.This study aims to explain the extent to which brain mapping patterns follow behavioral patterns of auditors and accountants' judgments when assessing evidence for decisions involving going concern. This multidisciplinary research involved investigating the relation between the theory of belief revision, neuroscience, and neuroaccounting with a sample of auditors and accountants. We developed a randomized controlled trial study with 12 auditors and 13 accountants. Auditors and accountants presented similar judgments about going concern, specially demonstrating greater sensitivity to negative evidence. Despite similar judgments, results showed diverging brain processing patterns between groups, as distinct reasoning was used to reach going concern estimates. During the decision process, auditors presented homogeneous brain processing patterns, while accountants evidenced conflicts and greater cognitive effort. For both groups, the occurrence of maximization (minimization) of judgments is observed in brain areas associated with identification of needs and motivations linked to individuals' relations with their social group. This was strengthened by the lack of significant differences between the regression maps of auditors and accountants, leading to interpretation of the groups' findings as homogeneous brain behavior. Despite familiarity with the executed task and knowledge of auditing standards, as a result of the greater use of algorithmic reasoning the auditors' judgments were similar to that of accountants. On the other hand, the accountants' greater cognitive effort, due to the experiencing of greater conflict in the decision-making process, made them use more quantic brain processing abilities, which are responsible for conscious reasoning. This was observed in the maximizations (minimizations) of the estimates in brain areas related to concerns with the judgments' social repercussions, which culminated in some degree of "conservatism" in their decisions. Furthermore, these findings reveal another opportunity to discuss the assumption of the brain as the original accounting institution.
- Impact of Camera Deviation on Penile Curvature Assessment Using 2D Pictures(2018) NASCIMENTO, B.; CERQUEIRA, I.; MIRANDA, E. P.; BESSA, J.; IVANOVIC, R. F.; GUGLIELMETTI, G.; NAHAS, W.C.; SROUGI, M.; CHIESA, G. A. E.; CURY, J.© 2018 International Society for Sexual MedicineBackground: An accurate curvature assessment (CA) is required in the decision-making process for patients with Peyronie's disease. In-office CA following induced erection is the gold standard for CA, although penile photography is commonly used due to its convenience. Camera deviations during 2D image acquisition might affect CA accuracy. Aim: To investigate the impact of camera angle deviations on CA. Methods: 2D pictures were taken from 5 models with a known uniplanar curvature (40°, 45°, 60°, 90°, and 120°). The model was kept on a fixed point and the camera was rotated around it. Pictures were taken with every 10° increase in camera deviation from the optimal position. The camera rotated to a maximum of 90° deviation in both the vertical and horizontal planes. The pictures were analyzed by 2 different urologists using a goniometer. The expected apparent curvature (AC) and the corresponding picture assessment error (PAE = AC – real model curvature) were also calculated for each picture using trigonometry principles. Main Outcome Measure: Assessing PAE magnitude and patterns was our primary outcome. Secondary outcomes were intraobserver, interobserver, and observer-AC intraclass correlation coefficient (ICC). Results: 100 pictures were analyzed. Intraobserver reliability was high (ICC = 0.99) for both urologists. Interobserver and observer-AC correlation were also high (ICC = 0.996 and ICC = 0.992, respectively). When the camera rotated in the horizontal axis, the PAE underestimated the curvature for models with curvatures smaller than 90° and overestimated the reading of the 120° model. When the camera rotated in the vertical axis, PAE had an inverse effect. The PAE showed a tendency to increase exponentially with higher deviation, reaching almost 100% for a deviation of 80°. Nevertheless, analyzing its magnitude regardless of the curvature, PAE was always <5% for camera deviations of 0–20°. Clinical Implications: If using picture-based CA, clinicians should attempt to take a picture perpendicular to the curvature plane for the most accurate measurement in degrees. Many clinicians request that patients take 3 pictures in a standard fashion (craniocaudal, lateral, and frontal), and if this technique is to be used, an extra picture is recommended. Strength & Limitations: In our controlled environment, we were able to isolate CA errors due to camera angles from other confounders such as erection hardness. As a consequence, however, our results cannot be easily generalized. Conclusion: PAE due to non-optimal camera position is a complex phenomenon that affects CA depending on the rotation axis and the degree of penile curvature. Nevertheless, PAE is always <5% for camera deviations of 0–20°. Nascimento B, Cerqueira I, Miranda EP, et al. Impact of Camera Deviation on Penile Curvature Assessment Using 2D Pictures. J Sex Med 2018;15:1638–1644.
- Assessing distributed collaborative recommendations in different opportunistic network scenarios(2020-08-01) BARBOSA, L. N.; GEMMELL, J. F.; HORVATH, M.; HEIMFARTH, T.© 2020 Inderscience Enterprises Ltd.Mobile devices are common throughout the world, even in countries with limited internet access and even when natural disasters disrupt access to a centralised infrastructure. This access allows for the exchange of information at an incredible pace and across vast distances. However, this wealth of information can frustrate users as they become inundated with irrelevant or unwanted data. Recommender systems help to alleviate this burden. In this work, we propose a recommender system where users share information via an opportunistic network. Each device is responsible for gathering information from nearby users and computing its own recommendations. An exhaustive empirical evaluation was conducted on two different data sets. Scenarios with different node densities, velocities and data exchange parameters were simulated. Our results show that in a relatively short time when a sufficient number of users are present, an opportunistic distributed recommender system achieves results comparable to that of a centralised architecture.
- Evaluation of ROS Navigation Stack for Social Navigation in Simulated Environments(2021-08-05) PIMENTEL, F. A. M.; Plinio Thomaz Aquino Junior© 2021, The Author(s), under exclusive licence to Springer Nature B.V.Accuracy and safety are necessary characteristics in social navigation. These characteristics still constitute a challenge in this area. Yet, human comfort is the main goal in interactions involving human beings. The ROS Navigation Stack (RNS) allows the variation of local path planning methods. This paper consists in a comparative study of methods related to social navigation. This study promotes better social navigation on Home Environment Robot Assistant (HERA). This is a robot platform developed by FEI University Center. This work evaluated various parameter combinations: type of environments, types of obstacles, local and global planning algorithms and costmaps. The work also evaluated people in static, dynamic and interacting ways. This study observed aspects of safety, accuracy of estimated time and space. Other aspects observed are the smooth trajectory realized and respect for personal space. The experiments performed 1000 attempts for 37 combinations of methods, environments and sensors. In total, the experiments counted 37000 attempts. With these experiments, was possible to select a configuration for the navigation system. The point to the Timed Elastic Band (TEB) as a local planner and a proxemic costmap as a good combination. The results reach 97.6% of success in a more complex environment with this combination.