BPG-based compression of Poisson noisy images DOI
Богдан Коваленко, Vladimir Lukin

Опубликована: Окт. 13, 2023

The article is devoted to the lossy compression of noisy images, considering currently popular trend green computing. In this paper, we show how can be incorporated into solving an essential task image data transferring via communication lines and what benefits it bring area interest. We consider case images corrupted with signal-dependent Poisson-distributed noise, which quite common in real-life situations. Compression has several peculiarities, among there existence optimal operation point (OOP), a value control parameter (CCP) where compressed closer noise-free than original one. possible presence OOP for various visual quality metrics, both SSIM-based PSNR-based. recommendations its neighborhood are presented.

Язык: Английский

The Role of Image Processing and Deep Learning in IoT-Based Systems: A Comprehensive Review DOI

Isamadeen A. Khalifa,

Faris Keti

Deleted Journal, Год журнала: 2025, Номер 3(1), С. 165 - 179

Опубликована: Фев. 19, 2025

The rates at which IoT is expanding are tremendous, literally touching our daily life experiences through various applications such as smart city, healthcare, agriculture and industrial automation among-couple others. From amongst a number of diverse types data produced by devices, image has risen to the forefront one most useful tools for real-time identification decision making. critical contribution processing deep learning in improving systems discussed this paper. Image acquisition, preprocessing, segmentation feature extraction procedures form basis acquiring significant information from raw imagery data. approaches CNNs, RNNs, transfer learning, makes classification extraction, object detection more accurate fully automated. These technologies have been incorporated used traffic monitoring application, medical diagnosis, environmental monitoring, fault diagnosis industries. Nonetheless, issues resource availability, temporal delay security act barriers adoption microservices especially edges fogs computing. To overcome these constraints, enhancement on lightweight Learning, Edge AI privacy protection methodologies being advanced efficient, secure real time performance. Hence, trends federated 5G can also define future based systems. This paper systematically critically reviews recent advances towards application architectures providing insight into its profile, challenges trends. It meant guide researchers industry experts who working building smarter scalable efficient

Язык: Английский

Процитировано

0

Evaluation of Different Methods and Models for Grass Cereals’ Production Estimation: Case Study in Wheat DOI Creative Commons
F. Sala, Mihai Valentin Herbei

Agronomy, Год журнала: 2023, Номер 13(6), С. 1500 - 1500

Опубликована: Май 30, 2023

Adequate management of agricultural crops requires, among other things, accessible and sufficiently accurate methods for assessing plant nutrition crop vegetation status production estimation. Sustainable technologies are based on correct decisions, prompt interventions appropriate works, information in real time, the obtaining can be simple, accessible, relation to different user categories (e.g., farmers, researchers, decision makers). This study used mineral fertilization (NPK), with 11 experimental variants, ensure a controlled differentiated wheat plants, “Alex” cultivar. Regression analysis was obtain models estimating production, by on: (a) NPK fertilizers applied (F) variants; (b) physiological indices (PI), represented chlorophyll content (Chl), terms macroelement leaves, evaluated foliar diagnosis (Nfd, Pfd, Kfd); (c) imaging (IA) digital images calculated indices. A set obtained, precision levels statistical safety: R2 = 0.763, p 0.013 model (NPKF); 0.883, < 0.01 (NPKfd); 0.857, (Chl); 0.975, normalized rgb color parameters (RGB system); 0.925, DGCI index. The (F model) tested data, period six years, communicated studies. Fit degree between predicted yield F (six-year average) confirmed 0.717, compared 0.763 this study. obtained study, related cultivar, studies, but certain margin error, given coefficient values, specific equations. approach concept, methods, presented opportunities studies facilitate their comparative analysis, adaptation, and/or development form new that useful research, or practices, integration into strategies.

Язык: Английский

Процитировано

3

Application of Information Technology Law in India on IoT/IoE With Image Processing DOI
Ramesh C. Tripathi,

Poonam Gupta,

Rohit Anand

и другие.

Advances in computational intelligence and robotics book series, Год журнала: 2023, Номер unknown, С. 135 - 150

Опубликована: Июнь 30, 2023

In order to address the issues posed by emerging technologies like IoT, IP, and AI, legislation governing information technology in India is undergoing a period of fast development (AI). Since these are able gather, share, analyse massive quantities data, regulations necessary safeguard individuals' right privacy about such data. Companies organizations need have full awareness legal framework around IoT/IoE, image processing, AI guarantee that they compliance with laws been enacted. organisations may assist guaranteeing employing legally responsible way using knowledge gained through understanding consequences technologies. It has observed there several applications IT law on IoT/ IoE. Present research work considering role processing IoT environment.

Язык: Английский

Процитировано

3

A secure and robust color image watermarking method using SVD and GAT in the multiresolution DCHWT domain DOI Creative Commons
Boubakeur Latreche,

Hilal Naimi,

Slami Saadi

и другие.

The Journal of Engineering and Exact Sciences, Год журнала: 2023, Номер 9(10), С. 17317 - 01e

Опубликована: Ноя. 9, 2023

This paper introduces an innovative and robust watermarking technique for safeguarding the copyright of color digital images. The method operates within domain SVD-based multiresolution discrete cosine harmonic wavelet transforms. In this approach, pre-processing phase employs successive generalized Arnold transforms to encrypt RGB watermark layers, significantly enhancing security algorithm. Subsequently, blue layer host image undergoes R-level 2D-DCHWT processing. encrypted is embedded by altering singular values image's approximation coefficients. Additionally, a reliable extraction algorithm devised recover from potentially compromised watermarked images without requiring access original image. Through extensive experiments comparisons with other relevant algorithms, findings demonstrate that scheme effectively embeds into not only ensures high levels invisibility but also establishes robustness, making it well-suited protection copyrights.

Язык: Английский

Процитировано

2

A Comprehensive Analysis of Image Segmentation Using Advance Artificial Intelligence DOI
Nilotpal Pathak, Kiran Singh, Fadi Al‐Turjman

и другие.

Advances in Science, Technology & Innovation/Advances in science, technology & innovation, Год журнала: 2024, Номер unknown, С. 141 - 148

Опубликована: Янв. 1, 2024

Язык: Английский

Процитировано

0

A Research on Two-Stage Facial Occlusion Recognition Algorithm based on CNN DOI Open Access
Zhe Wang, Malathy Batumalay, Rajermani Thinakaran

и другие.

Engineering Technology & Applied Science Research, Год журнала: 2024, Номер 14(6), С. 18205 - 18212

Опубликована: Дек. 2, 2024

In recent years, pattern recognition has garnered widespread attention, especially in the domain of face recognition. Traditional methods have certain limitations unconstrained environments due to factors such as lighting, facial expressions, and poses. Deep learning can be used address these challenges. This paper proposes a comprehensive approach occlusion based on two-stage Convolutional Neural Network (CNN). Face verification aims at verifying whether two images belong same individual, it is more fundamental task compared The process essentially involves multiple instances verification, sequentially validating different individuals ultimately determine corresponding individual for each face. primary steps this research include detection, image preprocessing, landmark localization, extraction, feature matching recognition, 2D image-assisted 3D reconstruction. A novel CNN was designed detection alignment. first stage network dedicated search windows regressing vector boundaries. second utilizes assist reconstruction perform secondary cases not identified stage. method demonstrated excellent performance handling occlusions, achieving high accuracy datasets AFW FDDB. On test dataset, reached 97.3%, surpassing original 89.1%. outperforms traditional algorithms general approaches. study achieved efficient validation further unrecognized situations, contributing enhancement system performance.

Язык: Английский

Процитировано

0

IOT-BASED SOLUTION FOR DETECTION DEFECTS IN INFRASTRUCTURE OBJECTS USING RASPBERRY PI DOI Open Access

R. Mysiuk,

Volodymyr Yuzevych

Electronics and Information Technologies, Год журнала: 2023, Номер 21

Опубликована: Янв. 1, 2023

Monitoring the current condition of roads, materials and communication networks can often be difficult in accessibility damaged place to visually assess criticality condition. The state such infrastructure facilities makes it possible data collected from sensors, intermediate analysis is automated microcomputers or microprocessors. For a visual assessment condition, video with overlay filters was used highlight areas damage. first type object defect corrosion, which detected by applying color filter, second cracks key points filled an edge filter. transfer number pixels filtered frame Raspberry Pi performed using MQTT protocols Internet Things system. AWS IoT rule constantly reads logged messages subscription push necessary into Elasticsearch database for further analysis. Keywords : Thing, cloud technologies, defects, Pi, crack.

Язык: Английский

Процитировано

0

BPG-based compression of Poisson noisy images DOI
Богдан Коваленко, Vladimir Lukin

Опубликована: Окт. 13, 2023

The article is devoted to the lossy compression of noisy images, considering currently popular trend green computing. In this paper, we show how can be incorporated into solving an essential task image data transferring via communication lines and what benefits it bring area interest. We consider case images corrupted with signal-dependent Poisson-distributed noise, which quite common in real-life situations. Compression has several peculiarities, among there existence optimal operation point (OOP), a value control parameter (CCP) where compressed closer noise-free than original one. possible presence OOP for various visual quality metrics, both SSIM-based PSNR-based. recommendations its neighborhood are presented.

Язык: Английский

Процитировано

0