The identification of defects in apple leaf specimens is crucial for mitigating crop loss and maintaining harvest quality. This study investigates the applicability of an intensity detection simulation using an integrated optical cross-sectional modeling method for detecting defective apple leaf specimens. The technique utilizes customized 840 nm optical coherence tomography (OCT) as the imaging tool, visualizing sufficient depth with a micrometer resolution. Leaf specimens were collected from apple plantations in Korea and categorized as healthy, apparently healthy, and infected leaf specimens. The method involved using a peak-intensity detection technique to analyze OCT signal intensity variations in multi-layered leaf structures. The method enhances defect detection accuracy by precisely characterizing the optical properties of the leaf specimens. The results demonstrate the method's potential to identify morphological differences between leaf specimens from healthy and infected trees and, specifically, healthy leaf specimens from infected trees. Through the quantitative analysis of OCT images, including quantitative information on cross-sectional thickness and depth direction, the method provides valuable insights into the structural changes associated with leaf defects, such as discoloration, tissue degradation, and altered layer morphology. Implementing this method in apple orchards can lead to significant cost savings by enabling timely interventions to mitigate the impact of leaf defects on crop production.

Fusarium poae is a pathogen that is widespread in the temperate zone and poses a serious threat to crops due to its wide range of host plants (including cereals). Electronic nose measurements were performed on wheat grains infected with F. poae to evaluate the application of early detection of fungal infections. Wheat seeds were artificially inoculated to test the devices. Three, same-weight but different infection levels, variants of experiments were prepared: 3 g infected seeds with 12 g healthy seeds, 5 g infected seeds with 10 g healthy seeds, and 10 g infected seeds with only 5 g healthy seeds. The seeds were infected with fresh fragments of F. poae mycelium. Measurements were carried out for five constructive days, recording the changes in volatile odor compounds released each day. A custom-built, low-cost device based on Figaro Inc. TGS metal-oxide, semiconductor gas sensors, and commercially available PEN3 electronic nose device from Airsense Analytics GmbH was used for the experiment. A non-linear sensor response for measured sample odor was observed with both devices. Spoiled grain in a proportion of 1/15 of the sample could be detected by measuring the volatile components. However, the patterns of the sensor responses were different for various concentrations of spoiled grain in the measured samples.

Fourier Transform Infrared Spectroscopy (FT-IR) is a vibrational technique largely adopted for the study of bacterial biofilms [1]. FT-IR is a non-destructive method, that allows multiple analyses of the same biofilm, without damaging the bacteria [2, 3].

Pseudomonas aeruginosa represents a class of bacteria largely investigated since it is an opportunistic pathogen and it is now considered to be a primary infectious agent especially for its ability to form multi-resistant biofilms [4].

In the present investigation, we aimed to characterize P. aeruginosa biofilms grown on different substrates to define better experimental conditions more useful for investigating the interaction of these biofilms with external agents. In particular, we investigated biofilms grown on Teflon membranes, CaF₂ windows, and MirrIR slides (specific reflection FT-IR spectroscopy microscope slides). Different geometries were used for collecting spectra using the microscope stage of a Perkin Elmer Spectrum One spectrometer equipped with a mercury cadmium telluride (MCT) detector.

Multiple acquisitions of spectra were done, and statistical criteria were applied for monitoring and comparing them. The positive and negative aspects of the different examined substrates for biofilm formation and acquisition modes are presented and discussed.

References

[1] Amy R. Crisp, Bryn Short, Laurence Rowan, Gordon Ramage, Ihtesham U.R. Rehman, Robert D. Short, Craig Williams, Investigating the chemical pathway to the formation of a single biofilm using infrared spectroscopy. Biofilm, 6, 2023, 100141, doi.org/10.1016/j.bioflm.2023.100141.

[2] Alvarez-Ordonez, A, Mouwen DJM, Lopez M, Prieto M. J Microbiol Methods 2011; 84:369–78.

[3] Z. Pang, R. Raudonis, B.R. Glick, T. Lin, Z. Cheng, Antibiotic resistance in Pseudomonas aeruginosa: mechanisms and alternative therapeutic strategies, Biotechnol. Adv. [Internet]. 37 (1) (2019) 177–192. Available from: 10.1016/j.biotechadv.2018.11.013.

[4] Thi MTT, Wibowo D, Rehm BHA. Pseudomonas aeruginosa Biofilms. Int J Mol Sci. 2020 Nov 17;21(22):8671. doi: 10.3390/ijms21228671.

Dentin and enamel are the two main constituents of human teeth, and the detailed characterization of their biochemical properties is of fundamental relevance in many fields of dentistry research. Vibrational spectroscopies such as Fourier Transform Infrared (FT-IR) and Raman spectroscopy can be adopted to obtain precise information before and after chemical or physical teeth treatments [1-4].

In the present work, the two above-mentioned spectroscopic techniques have been used for investigating dentin and enamel powders and few-mm thick disks cut from human molar teeth. The teeth were sectioned vertically with a diamond saw. After preparation, the samples were stored in a dry state. Before laser processing, they were rehydrated with distilled water for 24 hours to restore the normal fully hydrated state. FT-IR spectra were acquired on dried samples in micro-ATR mode in the 4000-650cm⁻¹ wavenumber range using the microscope stage of a Perkin-Elmer SpectrumOne spectrometer with an MCT detector. Raman spectra were collected in the 400-3200cm^-1 spectral range using a confocal Raman microscope (Horiba Xplora) with a Peltier-cooled CCD, a 50X objective, and a 785nm laser.

FT-IR and Raman spectra clearly show the contributions of different sample components. The spectra obtained from dentin and enamel powders evidence the differences due to their chemical composition. The spectra from human tooth disks present different characteristics depending on the region of the samples from which they were collected, thus enabling a spatial characterization of the samples themselves on different scales. These results confirm that vibrational spectroscopies allow a detailed characterization of hard dental tissues at the microscopic level.

References

[1]K.B.Beć, J. Grabska,C. W. Huck, Analytica Chimica Acta 1133, 150e177 (2020).

[2]I.Otel, Quantum Beam Sci. 7, 5 (2023).

[3]C.Diez, M.Á.Rojo, J.Martín-Gil, P.Martín-Ramos, M.Garrosa, D.Córdoba-Diaz, Minerals 12, 28 (2022).

[4]M.Anwar Alebrahim, C.Krafft, J.Popp, IOP Conf.Series:Materials Science and Engineering 92, 012014 (2015).

Digital non-metric cameras with high-resolution capabilities are being used for various domains such as digital heritage, artefact documentation, art conservation and engineering applications. In this study the novel approach is in the application of the combined use of close-range photogrammetry (CRP) and mapping techniques, to capture the depth of the mural digitally serving as a database for preservation and archiving of historic murals. The open hall next to the main sanctuary of the Virupaksha temple in Hampi, Karnataka, India, which is also a UNESCO World Heritage site, depicts cultural events on a mural-covered ceiling. A mirrorless Sony Alpha 7 III camera with a full frame 24MP CMOS sensor mounted with a 50mm and 24mm lens has been used to acquire digital photographs with an image size of 6000 × 6000 pixels. The suggested framework incorporates five main steps: data acquisition, colour correction, image mosaicking, orthorectification, and image filtering. Results show a high level of accuracy and precision attained during the image capture and processing steps. A comparative study was performed in which for a 24 mm lens orthoimage resulted in an image size of 9,131 x 14,910 and pixel size of 1.05mm whereas for a 50mm lens it produced 14,283 x 21,676 image size and a pixel size of 0.596 mm of the mural on the ceiling. This degree of high spatial resolution is essential for maintaining the digital documentation having fine details of the artwork as well as its historical context, subtleties, and painting techniques. The study's findings demonstrate the effectiveness of digital sensors with the close-range photogrammetry (CRP) technique as a useful method for recording and preserving historical ceiling murals.

According to recent studies by USA CDC, a notable proportion of elderly individuals experience falls each year, with approximately 20% of these fall events resulting in serious injuries such as fractures or head trauma. Given this statistic, detecting fall events is crucial for individuals who are elderly or are at risk of falls due to medical conditions.

Meanwhile, time-of-flight (ToF) sensors are increasingly utilized for human pose and gesture recognition. This paper explores the application of low-resolution (8*8) ToF sensors for detecting fall events in indoor environments (e.g., bathroom). We present a novel retrospective fall confirmation approach based on XGBoost that integrates fall postures data from distance snapshots and suspected fall trajectories. Our experiment results demonstrate strong detection performance, including accuracy and response time compared to traditional methods, highlighting the efficacy of leveraging history posture change process from stored sensor data alongside real-time ranging data judgement. Moreover, we explore and discuss the possibilities to use the low-resolution ToF sensor to realize the assessment of the seriousness of a fall event, facilitating timely medical assistance.

This work contributes to the research on applying advanced sensors and machine learning to elderly care and healthcare tasks and underscores the capability of low-resolution ToF sensors in monitoring human activity while respecting privacy concerns.

An oil spill at sea represents a catastrophic environmental event resulting from the release of oil into marine ecosystems. These incidents pose substantial risks to marine biodiversity, wildlife habitats, and coastal populations, often engendering enduring and widespread repercussions. Cleaning up oil spills is costly due to logistical challenges. Accurate measurement of spill characteristics like volume, thickness, and area of spill is crucial before deploying clean-up crews to optimize resource allocation and reduce expenses. The main objective of this research is to use computer vision to detect oil spills and estimate its thickness, helping in decision-making processes to clean up the spill area. A system architecture proposed in this study integrates a drone equipped with a camera and GSM module to inspect sea areas and capture images. These images are processed using a deployed computer vision segmentation model to detect oil spills and estimate oil thickness. Predicted results helps in decision-making via a dedicated application by applying predefined criteria to determine the thickness of the spill which further help in taking actions for removal of oil spills. The computer vision model developed in this research could detect and estimate oil thickness with a 94% accuracy. The proposed system in this study uses instance segmentation to detect and segment oil spills in drone footage. This computer vision-based approach accurately identifies and outlines oil spill areas, aiding in the selection of efficient cleanup strategies. Real-time monitoring and assessment capabilities enable quick decision-making and effective response measures.

Current research in home automation focuses on integrating emerging technologies like Internet of Things (IoT), and machine learning to create smart home solutions that offer enhanced convenience, efficiency, and security. Benefits include remote control of household devices, optimized energy usage through automated systems, and improved user experience with real-time monitoring and alerts. In this study a TinyML (Tiny Machine Learning) based keyword spotting machine learning model and system is proposed which enables voice-based home automation. The proposed system allows users to control household devices through voice commands with minimal computational resources and real-time performance. The main objective of this research is to develop TinyML model for resource constrained devices. The system enables home systems to efficiently recognize specific keywords or phrases by integrating voice control for enhanced user convenience and accessibility. In this research the different voice keywords of users of different age groups have been collected in home environment and trained using machine learning algorithm. An, IoT based system is then developed utilizing the TinyML model to recognize specific voice command and perform home automation tasks. The model has achieved 98% accuracy with F1 score of 1.00 and 92% recall. The quantized model uses Latency of 11 ms, 19.8K of RAM and 55.0 K of flash for keyword classification which is a best fit for any resource constraint devices. The proposed system demonstrates the viability of deploying a keyword spotting model for home automation on resource-constrained IoT devices. The research helps in building efficient and user-friendly smart home solutions, enhancing the accessibility and functionality of home automation systems.

Accurate and continuous location monitoring of athletes helps meeting health and safety requirements and supporting the infotainment needs of large marathon events with thousands of participants. Currently the tracking of individuals and groups of athletes at mass sports events is only possible to a limited extent, due to weight, size and cost constraints of the necessary devices. At marathon events, the usual infrastructure for timekeeping is RFID technology, which allows only precise tracking at huge intervals, with heuristic and interpolative algorithms to estimate runner positions in between the measuring points. Setting up RFID tracking stations on site is also material and labour intensive. We instead propose a continuous, real-time tracking solution, relying on LPWAN LoRa GPS trackers. Due to the large geographical area and urban space in which marathon events take place, the positioning of static gateways cannot provide complete and continuous coverage. This research article presents an implementation with multiple LoRa trackers and mobile LoRa gateways installed on vehicle escorts to assess coverage quality. The tracking data collected by a receiving LNS is stored in a database. Three experiments were conducted at three different official running events, a 10 km race, a half marathon and a marathon. The backdrop for 42.195 km was the official Vienna Marathon 2024 with more than 35,000 participants. The experimental results under these realistic conditions show the reception quality of this approach, e.g. during the marathon, received packets from LoRa gateways were at an average distance of about 136 metres (σ 157m) from the tracker with an average update rate of 49 seconds across all trackers, using DR3/SF9. At greater distances the quality decreases, although some outliers were received up to a distance of two kilometres. A possible prospect is that LPWAN may repeat the history of RFID by entering the mass sports market from the industry.

Introduction:
The integration of artificial intelligence (AI) sensors in healthcare has the potential to considerably improve patient monitoring, diagnosis, and treatment. However, the opaque nature of many AI systems known as black boxes poses significant challenges regarding transparency, explainability, and trustworthiness, which are critical factors in health delivery and management. This research aims to develop a framework for designing and deploying explainable and transparent AI sensors in healthcare.
Methods:
Through a comprehensive literature review and empirical analysis, we identify the key requirements and challenges associated with developing transparent and explainable AI (XAI) systems for healthcare applications. Our proposed approach combines interpretable machine learning models, human-AI interaction mechanisms, and ethical guidelines to ensure that AI sensor outputs are comprehensible, auditable, and aligned with clinical decision-making processes.
Results and Discussion:
Our proposed framework encompasses three core components: (1) an interpretable AI model architecture that leverages techniques such as attention mechanisms, symbolic reasoning, and rule-based systems to provide human-understandable explanations; (2) an interactive interface that facilitates effective communication and collaboration between healthcare professionals and AI systems, enabling seamless integration of AI insights into clinical workflows; and (3) a robust ethical and regulatory framework that addresses issues of bias, privacy, and accountability in the deployment of AI sensors in healthcare. Through case studies and simulations, we demonstrate the efficacy of our approach in enhancing transparency, explainability, and trust in AI-powered healthcare applications.
Conclusions:
The proposed framework contributes to the responsible development of AI technologies and paves the way for improved patient outcomes, informed decision-making, and increased public acceptance of AI in healthcare. By addressing the challenges of transparency and explainability, our research facilitates the safe and ethical adoption of AI sensors in healthcare.