Hardware Accelarated Visual Tracking Algorithms. A Systematic Literature Review

Many industrial applications need object recognition and tracking capabilities. The algorithms developed for those purposes are computationally expensive. Yet ,real time performance, high accuracy and small power consumption are essential measures of the system. When all these requirements are combined, hardware acceleration of these algorithms becomes a feasible solution. The purpose of this study is to analyze the current state of these hardware acceleration solutions, which algorithms have been implemented in hardware and what modifications have been done in order to adapt these algorithms to hardware.

Pretraining Convolutional Neural Networks for Visual Recognition

Convolutional Neural Networks (CNN) have become the state-of-the-art methods on many large scale visual recognition tasks. For a lot of practical applications, CNN architectures have a restrictive requirement: A huge amount of labeled data are needed for training. The idea of generative pretraining is to obtain initial weights of the network by training the network in a completely unsupervised way and then fine-tune the weights for the task at hand using supervised learning. In this thesis, a general introduction to Deep Neural Networks and algorithms are given and these methods are applied to classification tasks of handwritten digits and natural images for developing unsupervised feature learning. The goal of this thesis is to find out if the effect of pretraining is damped by recent practical advances in optimization and regularization of CNN. The experimental results show that pretraining is still a substantial regularizer, however, not a necessary step in training Convolutional Neural Networks with rectified activations. On handwritten digits, the proposed pretraining model achieved a classification accuracy comparable to the state-of-the-art methods.

Local and Global Feature Extraction for Invariant Object Recognition

Perceiving the world visually is a basic act for humans, but for computers it is still an unsolved problem. The variability present innatural environments is an obstacle for effective computer vision. The goal of invariant object recognition is to recognise objects in a digital image despite variations in, for example, pose, lighting or occlusion. In this study, invariant object recognition is considered from the viewpoint of feature extraction. Thedifferences between local and global features are studied with emphasis on Hough transform and Gabor filtering based feature extraction. The methods are examined with respect to four capabilities: generality, invariance, stability, and efficiency. Invariant features are presented using both Hough transform and Gabor filtering. A modified Hough transform technique is also presented where the distortion tolerance is increased by incorporating local information. In addition, methods for decreasing the computational costs of the Hough transform employing parallel processing and local information are introduced.

Similarity measure based classification

Luokittelujärjestelmää suunniteltaessa tarkoituksena on rakentaa systeemi, joka pystyy ratkaisemaan mahdollisimman tarkasti tutkittavan ongelma-alueen. Hahmontunnistuksessa tunnistusjärjestelmän ydin on luokitin. Luokittelun sovellusaluekenttä on varsin laaja. Luokitinta tarvitaan mm. hahmontunnistusjärjestelmissä, joista kuvankäsittely toimii hyvänä esimerkkinä. Myös lääketieteen parissa tarkkaa luokittelua tarvitaan paljon. Esimerkiksi potilaan oireiden diagnosointiin tarvitaan luokitin, joka pystyy mittaustuloksista päättelemään mahdollisimman tarkasti, onko potilaalla kyseinen oire vai ei. Väitöskirjassa on tehty similaarisuusmittoihin perustuva luokitin ja sen toimintaa on tarkasteltu mm. lääketieteen paristatulevilla data-aineistoilla, joissa luokittelutehtävänä on tunnistaa potilaan oireen laatu. Väitöskirjassa esitetyn luokittimen etuna on sen yksinkertainen rakenne, josta johtuen se on helppo tehdä sekä ymmärtää. Toinen etu on luokittimentarkkuus. Luokitin saadaan luokittelemaan useita eri ongelmia hyvin tarkasti. Tämä on tärkeää varsinkin lääketieteen parissa, missä jo pieni tarkkuuden parannus luokittelutuloksessa on erittäin tärkeää. Väitöskirjassa ontutkittu useita eri mittoja, joilla voidaan mitata samankaltaisuutta. Mitoille löytyy myös useita parametreja, joille voidaan etsiä juuri kyseiseen luokitteluongelmaan sopivat arvot. Tämä parametrien optimointi ongelma-alueeseen sopivaksi voidaan suorittaa mm. evoluutionääri- algoritmeja käyttäen. Kyseisessä työssä tähän on käytetty geneettistä algoritmia ja differentiaali-evoluutioalgoritmia. Luokittimen etuna on sen joustavuus. Ongelma-alueelle on helppo vaihtaa similaarisuusmitta, jos kyseinen mitta ei ole sopiva tutkittavaan ongelma-alueeseen. Myös eri mittojen parametrien optimointi voi parantaa tuloksia huomattavasti. Kun käytetään eri esikäsittelymenetelmiä ennen luokittelua, tuloksia pystytään parantamaan.

New kernel functions and learning methods for text and data mining

Recent advances in machine learning methods enable increasingly the automatic construction of various types of computer assisted methods that have been difficult or laborious to program by human experts. The tasks for which this kind of tools are needed arise in many areas, here especially in the fields of bioinformatics and natural language processing. The machine learning methods may not work satisfactorily if they are not appropriately tailored to the task in question. However, their learning performance can often be improved by taking advantage of deeper insight of the application domain or the learning problem at hand. This thesis considers developing kernel-based learning algorithms incorporating this kind of prior knowledge of the task in question in an advantageous way. Moreover, computationally efficient algorithms for training the learning machines for specific tasks are presented. In the context of kernel-based learning methods, the incorporation of prior knowledge is often done by designing appropriate kernel functions. Another well-known way is to develop cost functions that fit to the task under consideration. For disambiguation tasks in natural language, we develop kernel functions that take account of the positional information and the mutual similarities of words. It is shown that the use of this information significantly improves the disambiguation performance of the learning machine. Further, we design a new cost function that is better suitable for the task of information retrieval and for more general ranking problems than the cost functions designed for regression and classification. We also consider other applications of the kernel-based learning algorithms such as text categorization, and pattern recognition in differential display. We develop computationally efficient algorithms for training the considered learning machines with the proposed kernel functions. We also design a fast cross-validation algorithm for regularized least-squares type of learning algorithm. Further, an efficient version of the regularized least-squares algorithm that can be used together with the new cost function for preference learning and ranking tasks is proposed. In summary, we demonstrate that the incorporation of prior knowledge is possible and beneficial, and novel advanced kernels and cost functions can be used in algorithms efficiently.

Optical Character Recognition for Symbian OS

This Master's thesis addresses the design and implementation of the optical character recognition (OCR) system for a mobile device working on the Symbian operating system. The developed OCR system, named OCRCapriccio, emphasizes the modularity, effective extensibility and reuse. The system consists of two parts which are the graphical user interface and the OCR engine that was implemented as a plug-in. In fact, the plug-in includes two implementations of the OCR engine for enabling two types of recognition: the bitmap comparison based recognition and statistical recognition. The implementation results have shown that the approach based on bitmap comparison is more suitable for the Symbian environment because of its nature. Although the current implementation of bitmap comparison is lacking in accuracy, further development should be done in its direction. The biggest challenges of this work were related to developing an OCR scheme that would be suitable for Symbian OS Smartphones that have limited computational power and restricted resources.

New Distance Transforms for Gray Level Image Compression

This thesis deals with distance transforms which are a fundamental issue in image processing and computer vision. In this thesis, two new distance transforms for gray level images are presented. As a new application for distance transforms, they are applied to gray level image compression. The new distance transforms are both new extensions of the well known distance transform algorithm developed by Rosenfeld, Pfaltz and Lay. With some modification their algorithm which calculates a distance transform on binary images with a chosen kernel has been made to calculate a chessboard like distance transform with integer numbers (DTOCS) and a real value distance transform (EDTOCS) on gray level images. Both distance transforms, the DTOCS and EDTOCS, require only two passes over the graylevel image and are extremely simple to implement. Only two image buffers are needed: The original gray level image and the binary image which defines the region(s) of calculation. No other image buffers are needed even if more than one iteration round is performed. For large neighborhoods and complicated images the two pass distance algorithm has to be applied to the image more than once, typically 3 10 times. Different types of kernels can be adopted. It is important to notice that no other existing transform calculates the same kind of distance map as the DTOCS. All the other gray weighted distance function, GRAYMAT etc. algorithms find the minimum path joining two points by the smallest sum of gray levels or weighting the distance values directly by the gray levels in some manner. The DTOCS does not weight them that way. The DTOCS gives a weighted version of the chessboard distance map. The weights are not constant, but gray value differences of the original image. The difference between the DTOCS map and other distance transforms for gray level images is shown. The difference between the DTOCS and EDTOCS is that the EDTOCS calculates these gray level differences in a different way. It propagates local Euclidean distances inside a kernel. Analytical derivations of some results concerning the DTOCS and the EDTOCS are presented. Commonly distance transforms are used for feature extraction in pattern recognition and learning. Their use in image compression is very rare. This thesis introduces a new application area for distance transforms. Three new image compression algorithms based on the DTOCS and one based on the EDTOCS are presented. Control points, i.e. points that are considered fundamental for the reconstruction of the image, are selected from the gray level image using the DTOCS and the EDTOCS. The first group of methods select the maximas of the distance image to new control points and the second group of methods compare the DTOCS distance to binary image chessboard distance. The effect of applying threshold masks of different sizes along the threshold boundaries is studied. The time complexity of the compression algorithms is analyzed both analytically and experimentally. It is shown that the time complexity of the algorithms is independent of the number of control points, i.e. the compression ratio. Also a new morphological image decompression scheme is presented, the 8 kernels' method. Several decompressed images are presented. The best results are obtained using the Delaunay triangulation. The obtained image quality equals that of the DCT images with a 4 x 4

MCMC Analysis Of Classical Time Series Algorithms

Identification of order of an Autoregressive Moving Average Model (ARMA) by the usual graphical method is subjective. Hence, there is a need of developing a technique to identify the order without employing the graphical investigation of series autocorrelations. To avoid subjectivity, this thesis focuses on determining the order of the Autoregressive Moving Average Model using Reversible Jump Markov Chain Monte Carlo (RJMCMC). The RJMCMC selects the model from a set of the models suggested by better fitting, standard deviation errors and the frequency of accepted data. Together with deep analysis of the classical Box-Jenkins modeling methodology the integration with MCMC algorithms has been focused through parameter estimation and model fitting of ARMA models. This helps to verify how well the MCMC algorithms can treat the ARMA models, by comparing the results with graphical method. It has been seen that the MCMC produced better results than the classical time series approach.

Physical Realization of an Optical Automatic Target Recognition System

In the modern warfare there is an active development of a new trend connected with a robotic warfare. One of the critical elements of robotics warfare systems is an automatic target recognition system, allowing to recognize objects, based on the data received from sensors. This work considers aspects of optical realization of such a system by means of NIR target scanning at fixed wavelengths. An algorithm was designed, an experimental setup was built and samples of various modern gear and apparel materials were tested. For pattern testing the samples of actively arm engaged armies camouflages were chosen. Tests were performed both in clear atmosphere and in the artificial extremely humid and hot atmosphere to simulate field conditions.

Kahden merkkijonon pisimmän yhteisen alijonon ongelma ja sen ratkaiseminen

Tämä tutkielma kuuluu merkkijonoalgoritmiikan piiriin. Merkkijono S on merkkijonojen X[1..m] ja Y[1..n] yhteinen alijono, mikäli se voidaan muodostaa poistamalla X:stä 0..m ja Y:stä 0..n kappaletta merkkejä mielivaltaisista paikoista. Jos yksikään X:n ja Y:n yhteinen alijono ei ole S:ää pidempi, sanotaan, että S on X:n ja Y:n pisin yhteinen alijono (lyh. PYA). Tässä työssä keskitytään kahden merkkijonon PYAn ratkaisemiseen, mutta ongelma on yleistettävissä myös useammalle jonolle. PYA-ongelmalle on sovelluskohteita – paitsi tietojenkäsittelytieteen niin myös bioinformatiikan osa-alueilla. Tunnetuimpia niistä ovat tekstin ja kuvien tiivistäminen, tiedostojen versionhallinta, hahmontunnistus sekä DNA- ja proteiiniketjujen rakennetta vertaileva tutkimus. Ongelman ratkaisemisen tekee hankalaksi ratkaisualgoritmien riippuvuus syötejonojen useista eri parametreista. Näitä ovat syötejonojen pituuden lisäksi mm. syöttöaakkoston koko, syötteiden merkkijakauma, PYAn suhteellinen osuus lyhyemmän syötejonon pituudesta ja täsmäävien merkkiparien lukumäärä. Täten on vaikeaa kehittää algoritmia, joka toimisi tehokkaasti kaikille ongelman esiintymille. Tutkielman on määrä toimia yhtäältä käsikirjana, jossa esitellään ongelman peruskäsitteiden kuvauksen jälkeen jo aikaisemmin kehitettyjä tarkkoja PYAalgoritmeja. Niiden tarkastelu on ryhmitelty algoritmin toimintamallin mukaan joko rivi, korkeuskäyrä tai diagonaali kerrallaan sekä monisuuntaisesti prosessoiviin. Tarkkojen menetelmien lisäksi esitellään PYAn pituuden ylä- tai alarajan laskevia heuristisia menetelmiä, joiden laskemia tuloksia voidaan hyödyntää joko sellaisinaan tai ohjaamaan tarkan algoritmin suoritusta. Tämä osuus perustuu tutkimusryhmämme julkaisemiin artikkeleihin. Niissä käsitellään ensimmäistä kertaa heuristiikoilla tehostettuja tarkkoja menetelmiä. Toisaalta työ sisältää laajahkon empiirisen tutkimusosuuden, jonka tavoitteena on ollut tehostaa olemassa olevien tarkkojen algoritmien ajoaikaa ja muistinkäyttöä. Kyseiseen tavoitteeseen on pyritty ohjelmointiteknisesti esittelemällä algoritmien toimintamallia hyvin tukevia tietorakenteita ja rajoittamalla algoritmien suorittamaa tuloksetonta laskentaa parantamalla niiden kykyä havainnoida suorituksen aikana saavutettuja välituloksia ja hyödyntää niitä. Tutkielman johtopäätöksinä voidaan yleisesti todeta tarkkojen PYA-algoritmien heuristisen esiprosessoinnin lähes systemaattisesti pienentävän niiden suoritusaikaa ja erityisesti muistintarvetta. Lisäksi algoritmin käyttämällä tietorakenteella on ratkaiseva vaikutus laskennan tehokkuuteen: mitä paikallisempia haku- ja päivitysoperaatiot ovat, sitä tehokkaampaa algoritmin suorittama laskenta on.

Evaluation of Bubble Formation and Break Up in Suppression Pools by Using Pattern Recognition Methods

During a possible loss of coolant accident in BWRs, a large amount of steam will be released from the reactor pressure vessel to the suppression pool. Steam will be condensed into the suppression pool causing dynamic and structural loads to the pool. The formation and break up of bubbles can be measured by visual observation using a suitable pattern recognition algorithm. The aim of this study was to improve the preliminary pattern recognition algorithm, developed by Vesa Tanskanen in his doctoral dissertation, by using MATLAB. Video material from the PPOOLEX test facility, recorded during thermal stratification and mixing experiments, was used as a reference in the development of the algorithm. The developed algorithm consists of two parts: the pattern recognition of the bubbles and the analysis of recognized bubble images. The bubble recognition works well, but some errors will appear due to the complex structure of the pool. The results of the image analysis were reasonable. The volume and the surface area of the bubbles were not evaluated. Chugging frequencies calculated by using FFT fitted well into the results of oscillation frequencies measured in the experiments. The pattern recognition algorithm works in the conditions it is designed for. If the measurement configuration will be changed, some modifications have to be done. Numerous improvements are proposed for the future 3D equipment.

Scalable algorithms for height field illumination

Global illumination algorithms are at the center of realistic image synthesis and account for non-trivial light transport and occlusion within scenes, such as indirect illumination, ambient occlusion, and environment lighting. Their computationally most difficult part is determining light source visibility at each visible scene point. Height fields, on the other hand, constitute an important special case of geometry and are mainly used to describe certain types of objects such as terrains and to map detailed geometry onto object surfaces. The geometry of an entire scene can also be approximated by treating the distance values of its camera projection as a screen-space height field. In order to shadow height fields from environment lights a horizon map is usually used to occlude incident light. We reduce the per-receiver time complexity of generating the horizon map on N N height fields from O(N) of the previous work to O(1) by using an algorithm that incrementally traverses the height field and reuses the information already gathered along the path of traversal. We also propose an accurate method to integrate the incident light within the limits given by the horizon map. Indirect illumination in height fields requires information about which other points are visible to each height field point. We present an algorithm to determine this intervisibility in a time complexity that matches the space complexity of the produced visibility information, which is in contrast to previous methods which scale in the height field size. As a result the amount of computation is reduced by two orders of magnitude in common use cases. Screen-space ambient obscurance methods approximate ambient obscurance from the depth bu er geometry and have been widely adopted by contemporary real-time applications. They work by sampling the screen-space geometry around each receiver point but have been previously limited to near- field effects because sampling a large radius quickly exceeds the render time budget. We present an algorithm that reduces the quadratic per-pixel complexity of previous methods to a linear complexity by line sweeping over the depth bu er and maintaining an internal representation of the processed geometry from which occluders can be efficiently queried. Another algorithm is presented to determine ambient obscurance from the entire depth bu er at each screen pixel. The algorithm scans the depth bu er in a quick pre-pass and locates important features in it, which are then used to evaluate the ambient obscurance integral accurately. We also propose an evaluation of the integral such that results within a few percent of the ray traced screen-space reference are obtained at real-time render times.