CyborGlogs are lifelong log files of personal experiences that are captured without conscious thought or effort. By creating cyborglogs on a continuous basis, we can enable various novel applications where the lifelong records are used as memory aids or for personal safety. This thesis presents the development of a mobile application that provides the tools to instantly capture, archive, recall, and share our personal experiences on widely available cameraphones. To achieve this goal, a client-server computational framework is designed and implemented to support real-time interaction among the users. Three fully functional prototypes supporting three major mobile platforms (J2ME, Symbian, and iPhone) are presented to show the feasibility and flexibility of this framework. Finally, this thesis shows an early prototype which demonstrates the idea of mediated reality on cameraphones using various integrated sensors. This prototype explores the possibilities of developing truly intelligent wearable applications on mobile devices in the future.

CyborGlogs are lifelong log files of personal experiences that are captured without conscious thought or effort. By creating cyborglogs on a continuous basis, we can enable various novel applications where the lifelong records are used as memory aids or for personal safety. This thesis presents the development of a mobile application that provides the tools to instantly capture, archive, recall, and share our personal experiences on widely available cameraphones. To achieve this goal, a client-server computational framework is designed and implemented to support real-time interaction among the users. Three fully functional prototypes supporting three major mobile platforms (J2ME, Symbian, and iPhone) are presented to show the feasibility and flexibility of this framework. Finally, this thesis shows an early prototype which demonstrates the idea of mediated reality on cameraphones using various integrated sensors. This prototype explores the possibilities of developing truly intelligent wearable applications on mobile devices in the future.

The aim of this thesis is to reconstruct 3D representation of underwater scenes from uncalibrated video sequences. Underwater visualization is important for underwater Remotely Operated Vehicles and underwater is a complex structured environment because of inhomogeneous light absorption and light scattering by the environment. These factors make 3D reconstruction in underwater more challenging. The reconstruction consists of the following stages: Image enhancement, feature detection and matching, fundamental matrix estimation, auto-calibration, recovery of extrinsic parameters, rectification, stereo matching and triangulation. For image enhancement, a pre-processing filter is used to remove the effects of water and to enhance the images. Two feature extraction methods are examined: 1. Difference of Gaussian with SIFT feature descriptor, 2. Harris Corner Detector with grey level around the feature point. Matching is performed by finding similarities of SIFT features and by finding correlated grey levels respectively for each feature extraction method. The results show that SIFT performs better than Harris with grey level information. RANSAC method with normalized 8-point algorithm is used to estimate fundamental matrix and to reject outliers. Because of the difficulties of calibrating the cameras in underwater, auto-calibration process is examined. Rectification is also performed since it provides epipolar lines coincide with image scan lines which is helpful to stereo matching algorithms. The Graph-Cut stereo matching algorithm is used to compute corresponding pixel of each pixel in the stereo image pair. For the last stage triangulation is used to compute 3D points from the corresponding pixel pairs.

There is an increasing trend towards object detection from aerial and satellite images. Most of the widely used object detection algorithms are based on local features. In such an approach, first, the local features are detected and described in an image, then a representation of the images are formed using these local features for supervised learning and these representations are used during classification . In this thesis, Harris and SIFT algorithms are used as local feature detector and SIFT approach is used as a local feature descriptor. Using these tools, Bag of Visual Words algorithm is examined in order to represent an image by the help of histograms of visual words. Finally, SVM classifier is trained by using positive and negative samples from a training set. In addition to the classical bag of visual words approach, two novel extensions are also proposed. As the first case, the visual words are weighted proportional to their importance of belonging to positive samples. The important features are basically the features occurring more in the object and less in the background. Secondly, a principal component analysis after forming the histograms is processed in order to remove the undesired redundancy and noise in the data, reduce the dimension of the data to yield better classifying performance. Based on the test results, it could be argued that the proposed approach is capable to detecting a number of geo-spatial objects, such as airplane or ships, for a reasonable performance.

Visual surveillance in the maritime domain has been explored for more than a decade. Although it has produced a number of working systems and resulted in a mature technology, surveillance has been restricted to the port facilities or areas close to the coastline assuming a fixed-camera scenario. This dissertation presents several contributions in the domain of maritime surveillance. First, a novel algorithm for open-sea visual maritime surveillance is introduced. We explore a challenging situation with a camera mounted on a buoy or other floating platform. The developed algorithm detects, localizes, and tracks ships in the field of view of the camera. Specifically, our method is uniquely designed to handle a rapidly moving camera. Its performance is robust in the presence of a random relatively-large camera motion. In the context of ship detection, a new horizon detection scheme for a complex maritime domain is also developed. Second, the performance of the ship detection algorithm is evaluated on a dataset of 55,000 images. Accuracy of detection of up to 88% of ships is achieved. Lastly, we consider the topic of detection of the vanishing line of the ocean surface plane as a way to estimate the horizon in difficult situations. This allows extension of the ship-detection algorithm to beyond open-sea scenarios.

Objektų klasterizavimas – tai viena iš duomenų gavybos (angl. data mining) sričių. Šių algoritmų pagrindinis privalumas – gebėjimas atpažinti grupavimo struktūrą be jokios išankstinės informacijos. Magistriniame darbe yra pristatomas vaizdų klasterizavimo algoritmas, naudojantis savaime susitvarkančius neuroninius tinklus (angl. Self-Organizing Map). Darbe analizuojami vaizdų apdorojimo, ypatingųjų taškų radimo bei palyginimo metodai. Nustatyta, kad SIFT (angl. Scale Invariant Feature Transform) ypatingųjų taškų radimas bei aprašymas veikia patikimiausiai, todėl būtent SIFT taškiniai požymiai yra naudojami klasterizavime. Darbe taip pat analizuojamas atstumo tarp paveikslėlių radimo algoritmas, tiriami skirtingi jo parametrai. Algoritmų palyginimui yra naudojamos ROC (angl. Receiver Operating Characteristic) kreivės ir EER (angl. Equal Error Rate) rodiklis. Vaizdų klasterizavimui yra naudojamas ESOM (Emergent Self-Organizing Map) neuroninis tinklas, jis vizualizuojamas U-Matrix (angl. Unified distance Matrix) pagalba ir tinklo neuronai skirstomi į klasterius vandenskyros algoritmu su skirtingu aukščio parinkimu. Magistriniame darbe demonstruojami klasterizavimo rezultatai su pavyzdinėmis paveikslėlių duomenų bazėmis bei realiais gyvenimiškais vaizdais. / Clustering algorithms – a field of data mining – aims at finding a grouping structure in the input data without any a-priori information. The master thesis is dedicated for image processing and clustering algorithms. There are point-feature detection, description and comparison methods analyzed in this paper. The SIFT (Scale Invariant Feature Transform) by D. Lowe has been shown to behave better than the other ones; hence it has been used for image to image distance calculation and undirectly in clustering phase. Finding distances between images is not a trivial task and it also has been analysed in this thesis. Several methods have been compared using ROC (Receiver Operating Curve) and EER measurements. Image clustering process is described as: (1) training of ESOM (Emergent Self-Organizing Map), (2) its visualization in U-Matrix, (3) neuron clustering using waterflood algorithm, and (4) image grouping according to their best-matching unit neurons. The paper demonstrates the image clustering algorithm on public object image databases and real life images from the Internet as well.

This study reports on the quantitative evaluation of a set of state-of-the-art feature detectors and descriptors in the context of repeat photography. Unlike most related work, the proposed study assesses the performance of feature detectors when intra-pair variations are uncontrolled and due to a variety of factors (landscape change, weather conditions, different acquisition sensors). There is no systematic way to model the factors inducing image change. The proposed evaluation is performed in the context of image matching, i.e. in conjunction with a descriptor and matching strategy. Thus, beyond just comparing the performance of these detectors and descriptors, we also examine the feasibility of feature-based matching on repeat photography. Our dataset consists of a set of repeat and historic images pairs that are representative for the database created by the Mountain Legacy Project www.mountainlegacy.ca. / Graduate

This thesis is focused on comparing methods for similarity image retrieval. Common techniques and testing sets are introduced. The testing sets are there to measure the accuracy of the searching systems based on similarity image retrieval. Measurements are done on those models which are implemented on the basis of presented techniques. These measurements examine their results depending on the input data, used components and parameters settings, especially the impact of image resolution on the retrieval precision is examined. These results are analysed and the models are compared. Powered by TCPDF (www.tcpdf.org)

This thesis focuses on bare hand gesture recognition by proposing a new architecture to solve the problem of real-time vision-based hand detection, tracking, and gesture recognition for interaction with an application via hand gestures. The first stage of our system allows detecting and tracking a bare hand in a cluttered background using face subtraction, skin detection and contour comparison. The second stage allows recognizing hand gestures using bag-of-features and multi-class Support Vector Machine (SVM) algorithms. Finally, a grammar has been developed to generate gesture commands for application control. Our hand gesture recognition system consists of two steps: offline training and online testing. In the training stage, after extracting the keypoints for every training image using the Scale Invariance Feature Transform (SIFT), a vector quantization technique will map keypoints from every training image into a unified dimensional histogram vector (bag-of-words) after K-means clustering. This histogram is treated as an input vector for a multi-class SVM to build the classifier. In the testing stage, for every frame captured from a webcam, the hand is detected using my algorithm. Then, the keypoints are extracted for every small image that contains the detected hand posture and fed into the cluster model to map them into a bag-of-words vector, which is fed into the multi-class SVM classifier to recognize the hand gesture. Another hand gesture recognition system was proposed using Principle Components Analysis (PCA). The most eigenvectors and weights of training images are determined. In the testing stage, the hand posture is detected for every frame using my algorithm. Then, the small image that contains the detected hand is projected onto the most eigenvectors of training images to form its test weights. Finally, the minimum Euclidean distance is determined among the test weights and the training weights of each training image to recognize the hand gesture. Two application of gesture-based interaction with a 3D gaming virtual environment were implemented. The exertion videogame makes use of a stationary bicycle as one of the main inputs for game playing. The user can control and direct left-right movement and shooting actions in the game by a set of hand gesture commands, while in the second game, the user can control and direct a helicopter over the city by a set of hand gesture commands.

In this thesis deep convolutional neural networks models and feature descriptor models are compared. Feature descriptors are paired with suitable chosen classifier. These models are a part of machine learning therefore machine learning types are described in this thesis. Further these chosen models are described, and their basics and problems are explained. Hardware and software used for tests is listed and then test results and results summary is listed. Then comparison based on the validation accuracy and training time of these said models is done.