Examensarbeten för masterexamen // Master Theses

Länka till denna samling:

https://odr.chalmers.se/handle/20.500.12380/29

Browse

Visar 1 - 20 av 1803

2D Finger Motion Tracking, Implementation for Android Based Smartphones
(2012) Khosravi Nahouji, Maryam; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
Human computer interaction grows day by day and nowadays it has been entered many aspects of the life. Through this growth it has covered not only the desktop and laptop PCs but also mobile phones and specifically Smartphones. Therefore it worth to consider this topic and try to make a new functionality for these kinds of interactions. The intention is to find a practical way to make a new use of the video camera of the Android based Smartphones and track a single finger using the camera as sensor. To track an object in Android platform it needs some special prerequisite. It means that there should be a way to make the tracking applicable on Smartphones. for this purpose the solution this work has come up with is to use OpenCV library which is an open source library specialized for computer vision applications. This library is applicable on Android platform and therefore can be considered for our intention. Also considering different methods applied for hand tracking purpose in earlier works which has been studied in this work, the intention was to implement a method which was not been used by later studied works. Therefore the method of tracking used in this work is called “Motion Templates”. This method is a method represented by OpenCV library and used in this work. To reduce the complexity of the work the first mile-stone which is implemented in this work is to track the finger on PC and then try to port the method and algorithm to Android platform. The tracking is done successfully on the PC using this method and it can be recommended as a possible way of tracking to be considered on Android platforms as a future work.
3D head scanner
(2013) Jedvert, Magnus; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
The advent of cheap depth cameras such as Microsoft Kinect together with modern reconstruction algorithms implemented for the Graphics Processing Unit (GPU) o ers the potential of many new exciting applications. In this thesis, a scanning booth equipped with three Kinect cameras is built, where a user can scan their head and upper body into a high-quality textured 3D model. This is done using a variant of the KinectFusion algorithm, adapted to work with multiple cameras. The system operates in real-time and the reconstructed model is presented within seconds.
3D Object Classification using Point Clouds and Deep Neural Network for Automotive Applications
(2019) Larsson, Christian; Norén, Erik; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
Object identification is a central part of autonomous cars and there are many sensors to help with this. One such sensor is the LIDAR which creates point clouds of the cars surrounding. This thesis evaluates a solution for object identification in 3D point clouds with the help of a neural network. A system named DELIS (DEtection in Lidar Systems), which takes a point cloud generated from a LIDAR as input, is designed. The system consists of two subsystems, one non-machine learning algorithm which segments the point cloud into clusters, one for each object, and a neural network that classifies this clusters. The final output is then the classes and the coordinates of the objects in the point cloud. The result of this thesis is a system named DELIS that can identify between pedestrians, cars, and cyclists.
3D Window Manager Prototype
(2009) Jonsson, Andreas E; Järbratt, Marcus; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
The standard graphical user interfaces on leading operating systems today all use a two dimensional approach for interaction and visualization. Handling several applications using this approach often leads to a cluttered work space which is hard to manage efficiently. Research within the 3D window management area has been performed and many products exist today which utilize 3D in one way or another, but none of these have had any impact on how windows are managed in leading operating systems. The main goal with this thesis was to implement a prototype which utilized 3D in order to manage open windows in a structured way. The result is a system where the user can interact with windows in a 3D world. These windows can be placed and grouped on certain designated areas, which the user is able to navigate between quickly along predefined paths. In order to work with windows they are brought up from the 3D environment to a conventional two dimensional work space. Different users that tried out the prototype encountered no big issues with how the prototype solved navigation and interaction with objects in a 3D environment. While some of them appreciated the product as a whole, some could not see how this would benefit them during their daily computer usage, although they saw advantages with some of the functionality. // Sammanfattning De grafiska användargränssnitt som används i de ledande operativsystemen idag använder ett två-dimensionellt tillvägagångssätt för interaktion och visualisering. Att hantera flera applikationer på detta sätt leder ofta till en rörig arbetsyta som är svår att arbeta med effektivt. Forskning inom fönsterhantering i 3D har utförts och många produkter finns i dagsläget som utnyttjar 3D på ett eller annat sätt, men ingen av dessa har haft något genomslag för hur fönster hanteras i de ledande operativsystemen. Huvudmålet med denna rapport var att implementera en prototyp som utnyttjade 3D för att hantera öppna fönster på ett strukturerat sätt. Resultatet är ett system där användaren kan interagera med fönster i en 3D-värld. Dessa fönster kan placeras och grupperas på särskilda ytor, som användaren snabbt kan navigera mellan längs förbestämda banor. För att arbeta med fönster tas de upp från 3Dmiljön till en konventionell två-dimensionell arbetsyta.En grupp användare som prövade prototypen stötte inte på några större hinder med hur prototypen löste navigering och interaktion med objekt i en 3D-miljö. Medan några av dem uppskattade produkten i sin helhet, hade en del svårt att se hur detta skulle gynna dem i sin dagliga datoranvändning, även om de såg fördelar med delar av funktionaliteten.
5G User Plane Load Simulator
(2021) DÜSTERDIECK, OLOF; HUDA, TASDIKUL; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Papatriantafilou, Marina; Duvignau, Romaric
As the commercial introduction of 5G networks is getting closer, the deployment of testing technology able to perform at the requirements specified for 5G, is of utmost importance. As a company on the leading edge of 5G deployment, Ericsson manages mobile traffic by developing a gateway known as the Evolved Packet Gateway (EPG). EPG is the main component responsible for bridging the gap for data packets between 5G base stations and servers on the Internet. Following the 5G design, EPG is split into a Control Plane (CP) and a User Plane (UP). The purpose of this thesis is to implement a 5G network simulator that is able to load test a physical 5G User Plane which is used to route traffic in a mobile network. The simulator is able to simulate traffic in both the uplink and downlink direction as well as sending traffic on several different user sessions. To simulate traffic, three different traffic models, steady-rate, step-wise rate and Poisson processes, are used. In each simulation, the traffic is stateless and follows a client-to-server architecture. All parts included in this architecture is simulated, except for the UP. To analyze the results of a test, we define several different performance metrics such as throughput and latency. These are evaluated using evaluation techniques such as Control Charts. Several sets of experiments are performed in which we verify the implemented load types and measure the maximum rate of both the simulator and the UP. Using these results, it is possible to analyze the scalability of the UP and our simulator. We conclude that our simulator performs above the required rate to load test the UP for most scenarios with differing number of users and flows per each user. Where this is not the case, ideas and suggestions on how the simulator can be enhanced additionally are given. We also conclude that the UP scales very well with an increased number of users and that it performs above the proposed 5G requirements.
A Battery Management Unit
(2010) Fisk, Heidi; Leijgård, Johan; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
Emissions from vehicles and decreasing oil resources are pressing matters in today’s society. Electric vehicles are considered, by some, to be the solution to those problems. For years the available batteries have been too heavy and inefficient to be used in commercial electric vehicles. This thesis considers a new cell technology, LiFePO4, which was investigated and implemented into a battery pack for a prototype electric go-cart. The LiFePO4 cells need to be supervised by an efficient Battery Management Unit to function. The Battery Management Unit implemented in this thesis was based on minimal hardware combined with a prototype board for active cell balancing. The results show that it is possible to build such a lightweight Battery Management Unit, but with the loss of accuracy in the system.
A Case Study for Progressive Algorithms - An Investigation into Progressive Extraction of Intermediate Solutions for The Weighted Interval Scheduling Problem
(2020) Ringmark, Johannes; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Johansson, Richard; Damaschke, Peter
Asserted convergence characteristics shared by all serial algorithms have traditionally not been as prominent a measure of quality as time complexity has been. In instances relying on the exploration of large datasets, a convergence bound on interactively guided exploration approaches could act complementary to traditional time complexity, and constitute a alternative starting point for algorithm design. In an attempt to further explore the plausibility of this conjecture, a theoretical framework (Alewijnse, 2014) for convergence analysis through decomposition into consecutive intermediate computations is adopted, and its resulting intermediate solutions are used as a mean of empirical algorithm convergence analysis and categorization. Different scenarios related to the weighted interval scheduling problem are explored in this light, which is chosen primarily based on its documented compatibility with dynamic programming approaches. The result is presented as asymptotic upper bound functions on the convergence along with conjectures on upper bound hardness with respect to two different error functions, one adapted for stochastically and one for deterministically rooted algorithms.
A Case Study of Feature Location in an Open Source Embedded System
(2018) Gu, Wanzi; Shen, Hui; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
In this master thesis, we conducted a case study to look for effective methodologies of feature location in software maintenance. We started with domain study and analysis on the target subject to obtain first hand knowledge and background information about the system. Later we came up with two methodologies that could be systematically applied to the chosen subject to identify and locate features. The methodologies were named as Release Log and Source Code methodology since these were the artifacts where the feature are identified from. The methods were applied to the system, and we found totally 44 features including both mandatory and optional features. Our methods could also be applied to other systems with similar development process. Finally we also measured the characteristics of the features found using some metrics, the results also showed positive correlations between several pairs of measurement metrics.
A Case Study of Interactive Conflict-Resolution Support in Software Configuration
(2016) Jonsson, Daniel; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
The Linux kernel is one of the largest highly configurable systems, with more than 13,000 configuration options. Although a description is provided for many of the configuration options, configuring the kernel has been identified as troublesome by both users and developers. To assist the user in satisfying unmet dependencies when configuring a system, an algorithm for resolving configuration conflicts called RangeFix has been proposed by academia. In this case study, we explore how RangeFix can be integrated with the Linux kernel configurator xconfig. We develop a prototype based on xconfig, where an existing Scala implementation of RangeFix is integrated to generate fixes and help the user resolve unmet dependencies. The workflow for configuring the kernel, supported by this prototype, is evaluated with Linux users through a survey. We find that the prototype is promising and can be useful in certain scenarios. We also evaluate the existing Scala implementation of RangeFix in terms of correctness and performance. Although the correctness is found to be good, it is much slower than the users’ expectation. How RangeFix can be implemented to comply with the Linux kernel community’s requirements is also explored in this case study. The result is a partial C implementation of RangeFix, based on a SAT solver. This C implementation is also evaluated in terms of correctness and performance. Even if our C implementation is incomplete, we find that this is a feasible way to implement the algorithm. When finished, it might perform well enough to provide interactive conflict-resolution support in the Linux kernel configurators.
A Case Study of the Challenges with Applying Machine Learning in Industry: A Software Engineering Perspective
(2019) Eksmo, Samuel; Liu, Hanyan; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Feldt, Robert; Hebig, Regina
Data science is a growing trend and the advancement in machine learning and AI have been creating headlines in recent years. This has sparked an interest, not just in traditional IT-industries but also in businesses such as manufacturing, medicine and retail. Numerous industries are seeing potential in making their business more data driven and seeks to implement these trending technologies but few people know of the challenges that comes with applying it. This thesis aims at identifying the challenges, bridging the gap and lowering the entry barrier for engineers and researcher to contribute in the field of applied machine learning. In this case study, we examine how software engineers, data scientists and researchers can structure their work in order to increase the success rate of ML projects. Through interviews and a practical implementation test we analyze the underlying key concept that could help in bridging this gap. We conclude that software engineers can support in some initial data science activities, that communication between different stakeholders is crucial to the success of projects and that simpler ML models might be preferable in projects with time restrictions.
A Case Study on Test Optimisation and Visualisation of Diversity Information
(2019) Kulaglic, Amar; Helsing, Jonathan; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Steghöfer, Jan-Philipp; Gomes de Oliveira Neto, Francisco
Continuous Integration (CI) pipelines are vital in the implementation of CI and in the feedback cycles that surround automated testing in CI environments. A feedback cycle represents the time it takes from performing a commit until test results are ready. A significant problem in CI and automated testing is the long feedback cycles that come due to the increasing size of the test repository when executing test suites. The increased length of test execution is what this thesis will address using test case prioritisation. Through a design science methodology, we developed a tool and evaluated it by performing a case study at Volvo Car Corporation. The case study consists of two parts: evaluating the visualisation of data usually hidden during prioritisation, and data gathering and statistical analysis related to the performance of different distance measures and test case data. We have identified that similarity maps and history plots are good visualisation to enhance test decision making and maintaining and improving test repositories. Moreover, we have discovered the potential of using previous executions of test cases to determine their similarity.
A chatbot-based graphical user interface for hospitalized patients: Empowering hospitalized patients with a patient-centered user experience and features for self-service in a bedside tablet
(2019) LEVIN, JOHAN; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Björk, Staffan; Torgersson, Olof
Hospitalized patients often feel left out and uninformed during their treatment. Published scientific research and statistics points at several factors where digitization is able to make way for a better and foremost patient-centered healthcare. In order to promote a healthcare that is considered patient-centered it is furthermore of importance to make them participants in their healing. The findings of a performed design process promotes the fact that it is not only vital to thorough understand their needs and expressed requests, but also provide features for comfort. Furthermore, six categories of features have been defined in: Accessibility, Amenities, Hospital Services, Modularity, Privacy and Treatment Services. Suggestions have been provided for each category and are motivated by published scientific research as well as findings from the performed design work.
A Cluster Analysis Framework using DiMaxL - Clustering and data reduction in the presence of noise
(2010) Werner, Philip; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
The goal of this project was to develop a framework which can be used to make accurate predictions on large, sampled, data sets where statistical outliers are present. A secondary aim was to develop a method to reduce the large amount of data sometimes available, but all of it not always useful when making a prediction. The framework was developed using the DiMaxL algorithms and was tested on data sets taken from biology. These data sets are protein measurements where a large amount of statistical outliers are present. The results indicate that the method can accurately detect patterns even in presence of large amount of noise without any excessive overfitting. In the case of data reduction, the accuracy of the method is more sensitive to the amount of available data, and a semi-automatic procedure is recommended. In conclusion, the framework developed, is able to effectively remove noise while detecting the underlying pattern present, even in complex correlations.
A Combinatorial Approach to Protein Mixture - Identification based on Mass Spectrometry Data
(2009) Nosratinia, Mohsen; Chalmers tekniska högskola / Institutionen för data- och informationsteknik, Datavetenskap (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering, Computing Science (Chalmers)
Protein mixture identification by Mass Spectrometry (MS) data using massmapping experiments has become a powerful method in identification, and in some cases quantification, of proteins in samples taken from tissues. The mainstream approaches for database search method usually impose an upper bound of 2 or 3 on the number of proteins in a sample mixture. In this thesis the problem in its most general from is reformulated as a hitting set problem over (finding transversals of) a hypergraph. The limitations and extent of applicability of this approach, considering different error sources in MS data are addressed.
A Comfortable Transition to the Electric – Terminal Tractor Investigating the Design Process of Developing a Dashboard Interface for an Electric Terminal Tractor
(2022) Andréasson, Filip; Boman, Christoffer; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Chalmers University of Technology / Department of Computer Science and Engineering; Dahlstedt, Palle; Wozniak, Paweł W.
With the world gradually shifting toward renewable energy sources, many fossildriven industrial vehicles must transition to electric counterparts. Such transitions can be imposing on the comfort of users and poses an interesting design challenge. In this thesis, we perform a user-centred design process to create a digital dashboard interface for a new electric terminal tractor. We investigate how this process affects the user experience of the final interface concept, and what to consider when designing for a comfortable transition from traditional to novel interfaces. The process was challenged by a low availability of users, resorting to alternative methods to accomplish user-centred design. The final concept was developed through three design iterations, using legacy bias as a tool for creating comfortable transitions. Each iteration resulted in a set of concepts which were subsequently evaluated. The final concept was well-received by all stakeholders, and it managed to improve the user experience for terminal tractor operators, while still remaining a comfortable transition from traditional interfaces.
A Comparative Study of Segmentation and Classification Methods for 3D Point Clouds
(2016) NYGREN, PATRIK; Jasinski, Michael; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
Active Safety has become an important part of the current automotive industry due to its proven potential in making driving more joyful and reducing number of accidents and causalities. Different sensors are used in the active safety systems to perceive the environment and implement driver assistance and collision avoidance systems. Light detection and ranging (LIDAR) sensors are among the commonly utilized sensors in these systems; a LIDAR produces a point cloud from the surrounding and can be used to detect and classify objects such as cars, pedestrians, etc. In this thesis, we perform a comparative study where several methods to both segment Region Growing and Euclidian Clustering) and classify (Support Vector Machines, Feed Forward Neural Networks, Random Forests and K-Nearest Neighbors) point clouds from an urban environment are evaluated. Data from the KITTI database is used to validate the methods which are implemented using the PCL and Shark library. We evaluate the performance of the classification methods on two different sets of developed features. Our experiments show that the best accuracy can be obtained using SVMs, which is around 96.3% on the considered data set with 7 different classes of objects.
A Comparative study of the Cache Coherence and Moving Computation to Data Approach
(2012) Biswas, Animesh; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
As multicore computers are gaining widespread use, the main challenge for the future is how to design multicore systems that scale to hundreds of processor cores. One of the problems is how to implement the so popular shared memory model efficiently for future many-core systems. A critical mechanism to realize shared memory is to use cache coherence which allows multiple copies of a single memory block to be distributed across the caches attached to each core. Unfortunately, it is challenging to scale cache coherence mechanisms to hundreds of cores due to the latencies associated with communicating values and due to the complexity of the mechanism. An interesting alternative that has been considered in a recent research project is to allow only a single copy of a shared data structure and to force the computations that manipulate that data structure to be executed on the core that owns the data structure. While some tentative ideas for how to envision such a systems has been isolated, it is not clear how to implement such a system efficiently and whether the performance will be competitive with standard cache coherence solutions. This thesis project aims at implementing a single-copy shared memory model on a Tilera system with 64 cores. In the project, a run-time system is designed and implemented, parallel applications are mapped to it and the performance of the system is established and compared with a system employing cache coherence. Single-copy memory model will perform better than the coherent-based shared memory model as long as shared data size is smaller than the primary cache size. Because, in the single-copy memory model, core that is responsible for modifying shared data will encounter cache hit each time after the first access. On the other hand, in the coherent-based shared memory model, cores have to bring shared data from the remote memory for each access. However, coherent-based approach will start performing better when shared data size will exceed the primary cache size. Because bringing data from the remote cache takes less CPU cycle than from the main memory. From the experimental result based on the critical sections of different applications (linear solver equation, Radiosity application of SPLASH-2 benchmark), I have found that single-copy memory model shows significant performance improvement over the coherent-based shared memory model. Because, in the single-copy memory model, core that modifies shared data structure encounters cache hit most of the time while executing critical section. Cache hit during accessing the shared data structures saves substantial amount of CPU cycles.
A comparison of database management systems DDoS attack robustness
(2022) Persgården, Jonathan; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Ali-Eldin Hassan, Ahmed; Duvignau, Romaric
A common issue for software companies is that the data size in their database grows larger than their current system can handle. To further complicate matters databases might be forced to process an increasing amount of queries, either due to a growing number of users or due to malicious parties stressing the database with DDoS attacks. One strategy to account for this is to replace a single server database with a database distributed over a cluster to allow scaling the databases’ resources to match varying requirements. This thesis project aims to provide measurements of the robustness during a DDoS attack of two different database management systems, DBMS. One is Microsoft SQL server usually used as a single server database and the distributed database MongoDb. This thesis will try to answer the questions "How is the performance of the DBMS affected by a DDoS attack of varying strength?" and "How does data size affect the DBMS performance?". This has been done by performing two test scenarios. One run on a single server where one Microsoft SQL server instance was compared to three MongoDb instances and one run on a cloud solution comparing two Microsoft SQL Server instances to two MongoDb instances. In both cases, the database instances were hosted on lightweight virtual machines, Docker containers. Results have been generated by measuring response times to these databases while various DDoS attack has been performed on them. The results show that both solutions work well for smaller data sizes while MongoDb is more when the data size grows. Making MongoDb a better choice if the data is expected to continuously grow.
A Compiler from CakeML to JavaScript
(2018) Nyberg, Oskar; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
This thesis presents a new compiler from CakeML to JavaScript with support for almost the entire CakeML language. In addition to the new compiler, a JavaScript syntax formalization has been defined together with formal semantics for a subset of JavaScript. The semantics include coverage for language features introduced as part of the ECMAScript 2015 standard. The new compiler, syntax formalization and semantics are implemented in the HOL4 theorem prover to allow for future verification of the new compiler. The new compiler enables CakeML programs to be run in web browsers on both desktop computers and smart phones and other contexts previously not available to CakeML.
A Complete Customer-to-Customer Shipment Scheduling Module Solving the CVRPTW and MSTPP problems
(2012) Hedenström, Marcus; Sjöblom, Mattias; Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers); Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
We implement a module for scheduling orders within a transportation network, considering both a pickup route between customers and a terminal,and the transportation between terminals. Our goal is to schedule the full set of orders, minimizing the total cost by having orders share transportation vehicles when possible, even in the non-trivial case when some orders need to be diverted from their own shortest path to do so. We evaluate the effectiveness of a genetic algorithm on the vehicle routing problem, and implement a serialization protocol.

Browse

Browsar Examensarbeten för masterexamen // Master Theses efter Titel

Sökresultat per sida

Sortera efter