The original serial implementation of the segmentation algorithm exhibits suboptimal performance, motivating a systematic optimization effort without compromising accuracy.To analyze the unoptimized serial code and apply iterative code and compiler optimizations while preserving segmentation accuracy.A multi‐stage optimization process was employed: Initial algorithm profiling to identify bottlenecks. Iterative application of static code transformations. Integration of compiler‐level optimizations at each stage.Performance testing demonstrates substantial speedups: 6x to 8x improvement over the original implementation, depending on optimization stage. Parallelization in the final step further enhances throughput without sacrificing correctness.Systematic, profile‐guided optimization by combining code refactoring and compiler tuning yields significant performance gains. The approach maintains accuracy while enabling efficient parallelization, offering a scalable template for optimizing similar compute‐intensive algorithms.

This study deals with the unfolded quantities based on the time intervals between successive neutron events from the aspect of the power law. Neutrons from spontaneous fission in special nuclear material induce fission in most cases. In this study, it was demonstrated that the inverse value of the number of induced neutrons decreases following a power function with the increasing radius of a plutonium metal sphere. In addition, it was considered an increase in the neutron background level with increasing altitudes. The inverse value of the mean neutron counting rate depending on altitudes can be described with a power function merely for higher elevations. A linear relationship was obtained by plotting the quantities on logarithmic axes against each other indicating generally a power law relationship for both investigated phenomena. The results of this study showed a connection between two seemingly unrelated neutron phenomena through power laws based on the distributions of time intervals between successive neutron counts. The empirical evidence implies that a connection between the observed quantities in a log-log plot is unchanged except for a multiplicative constant.

This paper presents a simulation study of a first order plus dead time system (FOPDT) controlled with proportional integral derivative (PID) controller tuned by optimizing several objective functions and treating several different scenarios. Objective functions used are Integral of Absolute Magnitude of Error (IAE), Integral of Time multiplied by the Absolute Error (ITAE), Integral of the Square Error (ISE), Integral of Time multiplied by the Square Error (ITSE) and Mean of the Integral Square Error (MISE). Scenarios include unit step response and total energy effort analysis: with and without disturbance present, unit step response in case of model to true system mismatch in time constants. Time constants being dominant time constant of the process and dead time constant. optimization algorithm used is particle swarm optimization (PSO) algorithm.

In this paper, we present an implementation and analysis of the mean shift algorithm. The mean shift is a general non-parametric mode finding/clustering procedure widely used in image processing and analysis and computer vision techniques such as image denoising, image segmentation, motion tracking, etc.

Artefacts caused by the presence of metallic implants and prosthesis appear as dark and bright streaks in computed tomography (CT) images, that obscure the information about underlying anatomical structures. These phenomena can severely degrade the image quality and hinder the correct diagnostic interpretation. Although many techniques for the reduction of metal artefacts have been proposed in literature, their effectiveness is still limited. In this paper, an application of a convolutional neural networks (CNN) to the problem of metal artefact reduction (MAR) in the image domain is investigated. Experimental results show that image-domain CNN can substantially suppresses streaking artefacts in the reconstructed images.