|
|
Insertion of 2D Graphics into a Scene Captured by a Stationary Camera
Nguyen, Son Hai ; Beran, Vítězslav (referee) ; Herout, Adam (advisor)
Augmented reality visualizes additional information in real-world environment. Main goal of this work is achieving natural looking of the inserted 2D graphics in a scene captured by a stationary camera with possibility of real time processing. Although several methods tackled foreground segmentation problem, many of them are not robust enough on diverse datasets. Modified background subtraction algorithm ViBe yields best visual results, but because of the nature of binary mask, edges of the segmented objects are coarse. In order to smooth edges, Global Sampling Matting is performed, this refinement greatly increased the perceptual quality of segmentation. Considering that the shadows are not classified by ViBe, artifacts were occurring after insertion of segmented objects on top of the graphics. This was solved by the proposed shadow segmentation, which was achieved by comparing the differences between brightness and gradients of the background model and the current frame. To remove plastic look of the inserted graphics, texture propagation has been proposed, that considers the local and mean brightness of the background. Segmentation algorithms and image matting algorithms are tested on various datasets. Resulted pipeline is demonstrated in various scenes.
|
|
|
Approximation of Sound Propagation by Neural Networks
Nguyen, Son Hai ; Bartl, Vojtěch (referee) ; Herout, Adam (advisor)
Za účelem nahrazení výpočtově náročných konvenčních numerických metod řešících diferenciální rovnice jsou neurální výpočty stále více prozkoumávány. Tato práce se zaměřuje na řešení časově nezávislé Helmholtzovi rovnice, která modeluje šíření ultrazvuku při transkraniální léčbě ultrazvukem. Při použití konvolučních neuronových sítí musí být data navzorkovaná na pravidelné mřížce, abychom odstranili dané omezení, navrhli jsme neurální výpočet založený na grafových neuronových sítích. Narozdíl od fyzikálně informovaných neuronových sítích (PINN) je potřeba náš model natrénovat pouze jednou, řešení pro množinu nových parametrů vyžaduje pouze dopředných chod. Model byl natrénovaný pomocí učení s učitelem, kde referenční data byly vypočítána pomocí konvenční metody k-Wave. Náš model má stabilní rozvinutí, přestože byl natrénovaný pouze s osmi iteracemi. Ačkoli byl model natrénovaný pouze na datech s jedním zdrojem vln, tak zvládne predikovat i vlnová pole s více zdroji i v mnohem větších výpočetních doménách. Náš model je schopen predikovat subpixelové body s větší přesností než lineární interpolace. Dále je naše řešení schopno predikovat vlnové pole i s podvzorkovaným Laplaciánem, kde jsou pouhé tři vzorky na jednu vlnovou délku. Nejsme si vědomi žádné existující metody fungující s takto řídkou diskretizací.
|
|
|
Detekce čárových kódů v obraze
Vašíček, Vojtěch ; Vaško, Marek (referee) ; Nguyen, Son Hai (advisor)
The aim of this work is to evaluate the current methods of barcode detection in images, test several selected methods and then design and implement improvements to some of these methods. In this work, the Scharr detection method, the OpenCV library detector and the YOLOv5 neural network were tested. The enhancement itself was performed by combining the YOLOv5 neural network and the OpenCV library barcode detector. The combined method was able to achieve an improvement in average precision at a threshold of 0.5 of 3.42% over the YOLOv5 method and 38.38% over the OpenCV library barcode detector.
|
|
|
Approximation of Sound Propagation by Neural Networks
Nguyen, Son Hai ; Bartl, Vojtěch (referee) ; Herout, Adam (advisor)
Za účelem nahrazení výpočtově náročných konvenčních numerických metod řešících diferenciální rovnice jsou neurální výpočty stále více prozkoumávány. Tato práce se zaměřuje na řešení časově nezávislé Helmholtzovi rovnice, která modeluje šíření ultrazvuku při transkraniální léčbě ultrazvukem. Při použití konvolučních neuronových sítí musí být data navzorkovaná na pravidelné mřížce, abychom odstranili dané omezení, navrhli jsme neurální výpočet založený na grafových neuronových sítích. Narozdíl od fyzikálně informovaných neuronových sítích (PINN) je potřeba náš model natrénovat pouze jednou, řešení pro množinu nových parametrů vyžaduje pouze dopředných chod. Model byl natrénovaný pomocí učení s učitelem, kde referenční data byly vypočítána pomocí konvenční metody k-Wave. Náš model má stabilní rozvinutí, přestože byl natrénovaný pouze s osmi iteracemi. Ačkoli byl model natrénovaný pouze na datech s jedním zdrojem vln, tak zvládne predikovat i vlnová pole s více zdroji i v mnohem větších výpočetních doménách. Náš model je schopen predikovat subpixelové body s větší přesností než lineární interpolace. Dále je naše řešení schopno predikovat vlnové pole i s podvzorkovaným Laplaciánem, kde jsou pouhé tři vzorky na jednu vlnovou délku. Nejsme si vědomi žádné existující metody fungující s takto řídkou diskretizací.
|
|
|
Constrained activity sequencing
Nguyen, Son Tung ; Barták, Roman (advisor) ; Čepek, Ondřej (referee)
Many scheduling problems can be seen as activity sequencing problems, where the activity sequence in demand satisfies certain constraints. A typical example is scheduling in the airline industry where the task is to assign to each aircraft a segment of flight and nonflight activities while guaranteeing certain required properties. This diploma thesis deals with such type of constrained activity sequencing. The aim is to propose a formal model of the problem, including specification of all constraints and objectives, capable of finding (near)optimal sequences of activities. The proposed model is based on Constraint programming techniques.
|
|
|
E-mail client for J2ME devices
Nguyen, Son Tung ; Machek, Pavel (advisor) ; Kosina, Jiří (referee)
The aim of the project is to create a optimized client for using on mobile J2ME devices capable of work with emails. Both POP3 and iMAP servers will be supported to retrieve mails. SMTP protocol will be used for sending emails. The client will also support MIME type attachments and display basic supported formats like: txt, jpg, png and other multimedia formats that are supported by user's mobile device. Standard mail boxes like Inbox, Outbox, Drafts, Trash... will be supported as well. Emails can be ordered by various user's criterias : by subjects, date, size and senders or recipients. The program will support many email accounts. Automatic new emails checking. Operations that take longer time can be halted and whose progress will be displayed. Inteligent T9-like addressbook.
|
|
|
Insertion of 2D Graphics into a Scene Captured by a Stationary Camera
Nguyen, Son Hai ; Beran, Vítězslav (referee) ; Herout, Adam (advisor)
Augmented reality visualizes additional information in real-world environment. Main goal of this work is achieving natural looking of the inserted 2D graphics in a scene captured by a stationary camera with possibility of real time processing. Although several methods tackled foreground segmentation problem, many of them are not robust enough on diverse datasets. Modified background subtraction algorithm ViBe yields best visual results, but because of the nature of binary mask, edges of the segmented objects are coarse. In order to smooth edges, Global Sampling Matting is performed, this refinement greatly increased the perceptual quality of segmentation. Considering that the shadows are not classified by ViBe, artifacts were occurring after insertion of segmented objects on top of the graphics. This was solved by the proposed shadow segmentation, which was achieved by comparing the differences between brightness and gradients of the background model and the current frame. To remove plastic look of the inserted graphics, texture propagation has been proposed, that considers the local and mean brightness of the background. Segmentation algorithms and image matting algorithms are tested on various datasets. Resulted pipeline is demonstrated in various scenes.
|
|
|
Constrained activity sequencing
Nguyen, Son Tung ; Barták, Roman (advisor) ; Čepek, Ondřej (referee)
Many scheduling problems can be seen as activity sequencing problems, where the activity sequence in demand satisfies certain constraints. A typical example is scheduling in the airline industry where the task is to assign to each aircraft a segment of flight and nonflight activities while guaranteeing certain required properties. This diploma thesis deals with such type of constrained activity sequencing. The aim is to propose a formal model of the problem, including specification of all constraints and objectives, capable of finding (near)optimal sequences of activities. The proposed model is based on Constraint programming techniques.
|
|
|
E-mail client for J2ME devices
Nguyen, Son Tung ; Machek, Pavel (advisor) ; Kosina, Jiří (referee)
The aim of the project is to create a optimized client for using on mobile J2ME devices capable of work with emails. Both POP3 and iMAP servers will be supported to retrieve mails. SMTP protocol will be used for sending emails. The client will also support MIME type attachments and display basic supported formats like: txt, jpg, png and other multimedia formats that are supported by user's mobile device. Standard mail boxes like Inbox, Outbox, Drafts, Trash... will be supported as well. Emails can be ordered by various user's criterias : by subjects, date, size and senders or recipients. The program will support many email accounts. Automatic new emails checking. Operations that take longer time can be halted and whose progress will be displayed. Inteligent T9-like addressbook.
|
|
|
The importance of Japanese investment in the Czech Republic: current status, trends and pespectives
Nguyen, Son ; Stuchlíková, Zuzana (advisor) ; Kačírková, Eliška (referee)
The aim of my thesis is to introduce the readers to the importance of Japanese investment in the Czech Republic as one of the most attractive countries in Central Europe. I have divided my thesis into 3 chapters. The first chapter is focused on the definition of foreign direct investment. I define their purpose and goals that investors want to achieve. In the second chapter I describe the development of Japanese investment in Central Europe, especially Poland, Slovakia and Hungary. In the last chapter, I describe Japanese investment in the Czech Republic. In thispart, i put emphasis on Japanese companies that are operating in the Czech Republic.
|
guest ::
RSS feed.