A review of various image fusion types and transforms

Utilizing multiple views of an image is an important approach in digital photography, video editing, and medical image fusion applications. Image fusion (ImF) methods are used to improve an image's quality and remove noise from the image signal, resulting in a higher signal-to-noise ratio. A complete assessment of the literature on the different transform kinds, techniques, and rules utilized in ImF is presented in this paper. To assess the outcomes, a white flower image was fused using discrete wavelet transform (DWT) and discrete cosine transform (DCT) techniques. For validation of results, the red, green, blue (RGB) and intensity hue saturation (IHS) values of individual and fused images were evaluated. The results obtained from the fused images with the spatial IHS transform method give a remarkable performance. Furthermore, the results of the performance evaluation using DWT and DCT fusion techniques show that the same peak signal to noise ratio (PSNR) of 114.04 was achieved for both PSNR 1 and PSNR 2 for DCT, and different results were obtained for DWT. For signal to noise ratio (SNR), SNR 1 and SNR 2 achieved slightly similar values of 114.00 and 114.01 for DCT, while a SNR of 113.28 and 112.26 was achieved for SNR 1 and SNR 2 respectively.

A Validation Roadmap of Multi-Physics Simulators of the Resonator of MW-Class CW Gyrotrons for Fusion Applications

For a few years the multi-physics modelling of the resonance cavity (resonator) of MW-class continuous-wave gyrotrons, to be employed for electron cyclotron heating and current drive in magnetic confinement fusion machines, has gained increasing interest. The rising target power of the gyrotrons, which drives progressively higher Ohmic losses to be removed from the resonator, together with the need for limiting the resonator deformation as much as possible, has put more emphasis on the thermal-hydraulic and thermo-mechanic modeling of the cavity. To cope with that, a multi-physics simulator has been developed in recent years in a shared effort between several European institutions (the Karlsruher Institut für Technologie and Politecnico di Torino, supported by Fusion for Energy). In this paper the current status of the tool calibration and validation is addressed, aiming at highlighting where any direct or indirect comparisons with experimental data are missing and suggesting a possible roadmap to fill that gap, taking advantage of forthcoming tests in Europe.

Grain Refinement and Improved Mechanical Properties of EUROFER97 by Thermo-Mechanical Treatments

EUROFER97 steel plates for nuclear fusion applications are usually manufactured by hot rolling and subsequent heat treatments: (1) austenitization at 980 °C for 30 min, (2) rapid cooling and (3) tempering at 760 °C for 90 min. An extended experimental campaign was carried out with the scope of improving the strength of the steel without a loss of ductility. Forty groups of samples were prepared by combining cold rolling with five cold reduction ratios (20, 40, 50, 60 and 80%) and heat treatments at eight different temperatures in the range 400–750 °C (steps of 50 °C). This work reports preliminary results regarding the microstructure and mechanical properties of all the cold-rolled samples and the effects of heat treatments on the samples deformed with the greater CR ratio (80%). The strength of deformed samples decreased as heat treatment temperature increased and the change was more pronounced in the samples cold-rolled with greater CR ratios. After heat treatments at temperature up to 600 °C yield stress (YS) and ultimate tensile strength (UTS) of samples deformed with CR ratio of 80% were significantly larger than those of standard EUROFER97 but ductility was lower. On the contrary, the treatment at 650 °C produced a fully recrystallized structure with sub-micrometric grains which guarantees higher strength and comparable ductility. The work demonstrated that EUROFER97 steel can be strengthened without compromising its ductility; the most effective process parameters will be identified by completing the analyses on all the prepared samples.

Engineering Interoperable, Plug-and-Play, Distributed, Robotic Control Systems for Futureproof Fusion Power Plants

Maintenance and inspection systems for future fusion power plants (e.g., STEP and DEMO) are expected to require the integration of hundreds of systems from multiple suppliers, with lifetime expectancies of several decades, where requirements evolve over time and obsolescence management is required. There are significant challenges associated with the integration, deployment, and maintenance of very large-scale robotic systems incorporating devices from multiple suppliers, where each may utilise bespoke, non-standardised control systems and interfaces. Additionally, the unstructured, experimental, or unknown operational conditions frequently result in new or changing system requirements, meaning extension and adaptation are necessary. Whilst existing control frameworks (e.g., ROS, OPC-UA) allow for the robust integration of complex robotic systems, they are not compatible with highly efficient maintenance and extension in the face of changing requirements and obsolescence issues over decades-long periods. We present the CorteX software framework as well as results showing its effectiveness in addressing the above issues, whilst being demonstrated through hardware that is representative of real-world fusion applications.

AI-Based Multi Sensor Fusion for Smart Decision Making: A Bi-Functional System for Single Sensor Evaluation in a Classification Task

Sensor fusion has gained a great deal of attention in recent years. It is used as an application tool in many different fields, especially the semiconductor, automotive, and medical industries. However, this field of research, regardless of the field of application, still presents different challenges concerning the choice of the sensors to be combined and the fusion architecture to be developed. To decrease application costs and engineering efforts, it is very important to analyze the sensors’ data beforehand once the application target is defined. This pre-analysis is a basic step to establish a working environment with fewer misclassification cases and high safety. One promising approach to do so is to analyze the system using deep neural networks. The disadvantages of this approach are mainly the required huge storage capacity, the big training effort, and that these networks are difficult to interpret. In this paper, we focus on developing a smart and interpretable bi-functional artificial intelligence (AI) system, which has to discriminate the combined data regarding predefined classes. Furthermore, the system can evaluate the single source signals used in the classification task. The evaluation here covers each sensor contribution and robustness. More precisely, we train a smart and interpretable prototype-based neural network, which learns automatically to weight the influence of the sensors for the classification decision. Moreover, the prototype-based classifier is equipped with a reject option to measure classification certainty. To validate our approach’s efficiency, we refer to different industrial sensor fusion applications.