Connectivity enforced bayesian superpixels

Download
2019
Eker, Onur
In this study a general flow for clustering-based superpixel (SP) extraction methods is presented, while each step is analyzed in detail, and improvements are proposed. Considering general SP extraction method steps, initial grid alternatives are examined. The necessity of initial grid refinement is studied and unlike current approaches, a novel Edge Based Refinement step which does not break regular grid structure is proposed. Label update constraints are also analyzed in terms of preserving regular initial tiling, and Just Connected method enforcing connectivity from the beginning is proposed. The requirement of adjusting one of hyper-parameters, iteration count, for different image resolutions and different number of SPs is eliminated by determining the number of iterations relative to SP area. Considering these proposals,extensions to the state-of-the-art SP methods, SLIC+ and LASP+, are proposed which normalize spatial term with SP spatial covariance. Novel cost-functions SLIC++ and LASP++ are also presented for a further improvement with the normalization of spectral term with SP specific dynamic parameter. Finally, a Bayesian classifier is proposed for pixels during SP label assignment. Based on improvements in various steps mentioned above, a family of superpixel extraction methods including, SLIC++/R, SLIC++/H, LASP++/R. LASP++/H, BSP/R and BSP/H are presented. For the evaluation, a novel Boundary Achievable Segmentation Accuracy metric is proposed that replaces three frequent metrics from the literature. Compactness and area under curve approaches are also proposed as evaluation methods to minimize any performance ambiguity for the literature benchmarks. Both proposed spectral term and spatial term improvements significantly increase accuracy of generated SPs with no execution-time burden. In addition, employing Bayesian classifier leads to generate more accurate SPs in a shorter amount of run-time. Besides, with the proposed label update criteria, connectedness of SPs are ensured during generation process that preserves regular grid topology enabling them to be fed into conventional neural-networks.

Suggestions

Limitations of three-phase Buckley-Leverett theory
Akın, Serhat (Informa UK Limited, 2005-07-01)
The broad objective of this study is to determine the limitations of the three-phase relative permeability estimation technique by using the experimental approach. In order to achieve this goal, three-phase relative permeability experiments were conducted on a Berea sandstone core plug by using brine, hexane and nitrogen gas. An unsteady-state analytical technique and a numerical technique where a black oil simulator was coupled with a global optimization algorithm in a least squares manner was used to comp...
Thermogravimetry of selected bentonites
Kök, Mustafa Verşan (Informa UK Limited, 2002-10-01)
In this research the application of thermal analysis techniques (TG-DTA) to the characterization of bentonite samples is presented. The thermal behavior of 4 bentonite samples was investigated using simultaneous thermogravimetry (TG/DTG) and differential thermal analysis (DTA) techniques. It was observed that all bentonite samples exhibit 2 weight loss events on heating. The first weight loss event corresponds to the loss of molecular water from the exchange layer, whereas the second weight loss event corre...
Modeling and implementation of local volatility surfaces in Bayesian framework
Animoku, Abdulwahab; Uğur, Ömür; Yolcu-Okur, Yeliz (2018-06-01)
In this study, we focus on the reconstruction of volatility surfaces via a Bayesian framework. Apart from classical methods, such as, parametric and non-parametric models, we study the Bayesian analysis of the (stochastically) parametrized volatility structure in Dupire local volatility model. We systematically develop and implement novel mathematical tools for handling the classical methods of constructing local volatility surfaces. The most critical limitation of the classical methods is obtaining negativ...
Statistical Approach of Two-three Parameters Rheological Models for Polymer Type Drilling Fluid Analysis
Kök, Mustafa Verşan (Informa UK Limited, 2010-01-01)
In this study, 45 KCl/polymer type of drilling fluid samples were prepared for statistical analysis. Six difference (600, 300, 200, 100, 6, and 3) readings at standard Fann viscometer speeds were determined in order to estimate the rheological model parameters of our different models, as follows: Bingham Plastic, Power-Law, Hershel-Bulkley, and Robertson-Stiff. Parameters of the rheological models are estimated by explicitly solving the required number of equations simultaneously. Different statistical tech...
Isoconversional methods to determine the kinetics of crude oils - thermogravimetry approach
Kök, Mustafa Verşan; Nurgaliev, Danis K. (Elsevier BV, 2018-08-01)
In this research, kinetics of four crude oils from different origins are determined by iso-conversional methods using thermogravimetry data (TGA-DTG). The experiments were performed at three different heating rates (5-10-15 K/min) between 300 and 1200 K. Thermal characteristics of the crude oil samples such as, reaction intervals and corresponding peak and burn-out temperatures are also determined. Three different iso-conversional methods, known as Starink, Kissinger and Friedman are used in order to determ...
Citation Formats
O. Eker, “Connectivity enforced bayesian superpixels,” Thesis (M.S.) -- Graduate School of Natural and Applied Sciences. Electrical and Electronics Engineering., Middle East Technical University, 2019.