Marco Wehrmeister

Dr. Marco Wehrmeister

Federal University of Technology - Parana (UTFPR) - Campus Curitiba
orcid

Research Focus & Metrics

Academic Career Length

23Years (2004-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 July 2018 in Sensors
Verified
Journal Article
Open access
PDF file
DODiulhio De Oliveira
Diulhio De Oliveira
MWMarco Wehrmeister
Marco Wehrmeister
The use of Unmanned Aerial Vehicles (UAV) has been increasing over the last few years in many sorts of applications due mainly to the decreasing cost of this technology. One can see the use of the UAV in several civilian applications such as surveillance and search and rescue. Automatic detection of pedestrians in aerial images is a challenging task. The computing vision system must deal with many sources of variability in the aerial images captured with the UAV, e.g., low-resolution images of pedestrians, images captured at distinct angles due to the degrees of freedom that a UAV can move, the camera platform possibly experiencing some instability while the UAV flies, among others. In this work, we created and evaluated different implementations of Pattern Recognition Systems (PRS) aiming at the automatic detection of pedestrians in aerial images captured with multirotor UAV. The main goal is to assess the feasibility and suitability of distinct PRS implementations running on top of low-cost computing platforms, e.g., single-board computers such as the Raspberry Pi or regular laptops without a GPU. For that, we used four machine learning techniques in the feature extraction and classification steps, namely Haar cascade, LBP cascade, HOG + SVM and Convolutional Neural Networks (CNN). In order to improve the system performance (especially the processing time) and also to decrease the rate of false alarms, we applied the Saliency Map (SM) and Thermal Image Processing (TIP) within the segmentation and detection steps of the PRS. The classification results show the CNN to be the best technique with 99.7% accuracy, followed by HOG + SVM with 92.3%. In situations of partial occlusion, the CNN showed 71.1% sensitivity, which can be considered a good result in comparison with the current state-of-the-art, since part of the original image data is missing. As demonstrated in the experiments, by combining TIP with CNN, the PRS can process more than two frames per second (fps), whereas the PRS that combines TIP with HOG + SVM was able to process 100 fps. It is important to mention that our experiments show that a trade-off analysis must be performed during the design of a pedestrian detection PRS. The faster implementations lead to a decrease in the PRS accuracy. For instance, by using HOG + SVM with TIP, the PRS presented the best performance results, but the obtained accuracy was 35 percentage points lower than the CNN. The obtained results indicate that the best detection technique (i.e., the CNN) requires more computational resources to decrease the PRS computation time. Therefore, this work shows and discusses the pros/cons of each technique and trade-off situations, and hence, one can use such an analysis to improve and tailor the design of a PRS to detect pedestrians in aerial images.

1 July 2010
Verified
Conference Paper
Closed access
TMTomas G. Moreira
Tomas G. Moreira
MWMarco Wehrmeister
Marco Wehrmeister
CPCarlos E. Pereira
Carlos E. Pereira
JPJean-Francois Petin
Jean-Francois Petin
ELEric Levrat
Eric Levrat
The design of modern embedded systems becomes more complex every day, due to the increasing amount of components and distinct functionalities incorporated into a single system. To address this situation, projects' abstraction level is continuously raised. In addition, techniques to speed up the code production process have emerged. These new trends lead to an important issue: the development of a consistent engineering process for systems co-design, from the requirements phase to the source code production. In this context, the UML is an interesting option for the embedded systems projects, leading to techniques to generate source code for both hardware and software. Currently, there are some works and commercial tools to generate source code from UML specifications to mainstream languages, such as C++ and Java. However, there are few works addressing the automatic source code generation for VHDL language, which is widely used in the embedded systems development. Thus, this work proposes a methodology to generate automatically VHDL source code from UML specifications. This methodology is supported by the GenERTiCA tool. A use case focused on the use of embedded systems for the maintenance systems is presented in this paper to demonstrate the feasibility of the proposed approach.

9 November 2020 in Sensors
Verified
Journal Article
Open access
PDF file
ACAlvaro Cantieri
Alvaro Cantieri
MFMatheus Ferraz
Matheus Ferraz
GBGuido Szekir Berger
Guido Szekir Berger
Marco Antonio Simoes TeixeiraMarco Antonio Simoes Teixeira
Marco Antonio Simoes Teixeira
Jose LimaJose Lima
Jose Lima
MWMarco Wehrmeister
Marco Wehrmeister
Realizing autonomous inspection, such as that of power distribution lines, through unmanned aerial vehicle (UAV) systems is a key research domain in robotics. In particular, the use of autonomous and semi-autonomous vehicles to execute the tasks of an inspection process can enhance the efficacy and safety of the operation; however, many technical problems, such as those pertaining to the precise positioning and path following of the vehicles, robust obstacle detection, and intelligent control, must be addressed. In this study, an innovative architecture involving an unmanned aircraft vehicle (UAV) and an unmanned ground vehicle (UGV) was examined for detailed inspections of power lines. In the proposed strategy, each vehicle provides its position information to the other, which ensures a safe inspection process. The results of real-world experiments indicate a satisfactory performance, thereby demonstrating the feasibility of the proposed approach.

1 May 2007
Verified
Conference Paper
Closed access
MWMarco Wehrmeister
Marco Wehrmeister
EFEdison P. Freitas
Edison P. Freitas
CPCarlos E. Pereira
Carlos E. Pereira
FWFlavio R. Wagner
Flavio R. Wagner
This work presents a proposal to use aspect orientation in the analysis and design of distributed embedded real-time systems (DERTS). These systems have several requirements directly related to their main characteristics, the so-called non-functional requirements (NFR), which refer to orthogonal properties, conditions, and restrictions that are spread out over the entire system. Pure object-oriented methods do not address successfully those so called cross-cutting concerns, so new technologies, like aspect orientation, are applied in order to fulfill this gap. The paper presents DERAF istributed embedded real-time aspects framework, an extensible and high-level framework (i.e. implementation-independent) to handle NFR of DERTS at earlier design stages. DERAF combines the use of aspects with RT-UML, aiming to separate the handling of non-functional from functional requirements in the model driven design of DERTS. The paper presents the use of DERAF on a case study of an unmanned air vehicle (UAV)

15 January 2013 in IEEE Transactions on Industrial Informatics
Verified
Journal Article
Closed access
MWMarco Wehrmeister
Marco Wehrmeister
CPCarlos Eduardo Pereira
Carlos Eduardo Pereira
FRFranz J. Rammig
Franz J. Rammig
Automation and control systems include many “intelligent” automation devices, which are usually implemented as complex embedded systems. New methods and tools are demanded to cope with the increasing design complexity, while keeping the project on schedule. Proper handling of nonfunctional system requirements is a key factor during the design of industrial automation systems, since in some application domains they are as important as (sometimes, more important than) functional requirements. This paper presents a model-driven engineering approach, which combines Unified Modeling Language (UML) and aspect-oriented software development (AOSD) to design real-time and embedded automation systems. The proposed approach allows a smooth transition from the initial phases to implementation by using software tools, comprising the system specification and the automatic generation of source code. By combining UML with model-level aspects and a script-base code generation tool, it enables the use of AOSD during system design and implementation, even though the target platform does not natively support such concepts. Experimental results on using this approach to design real-world examples of automation systems are presented. The results indicate a positive impact on the design of automation systems. The encapsulation of nonfunctional requirements was improved, increasing the reuse of developed artifacts. Generated source code statistics indicate that the proposed approach can generated a fair amount of code per model element.

1 October 2014 in Mechatronics
Verified
Journal Article
Closed access
MWMarco Wehrmeister
Marco Wehrmeister
Edison  Pignaton de FreitasEdison Pignaton de Freitas
Edison Pignaton de Freitas
ABAlécio Pedro Delazari Binotto
Alécio Pedro Delazari Binotto
CPCarlos Eduardo Pereira
Carlos Eduardo Pereira