Real Time Adaptive Approach for Image Processing Using Mobile Nodes

Darweesh, Ahmed Adel; Abouelfarag, Ahmed; Kadry, Rania

doi:10.1109/w-ficloud.2018.00031

Cited by 3 publications

(1 citation statement)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The initial sections of the paper describe the hardware and software setup for pre-processing of these images and determination of weight and reason for generation of potato waste. The latter sections discuss the experimental results and potential implementation within other food manufacturing applications [23][24][25].…”

Section: Introductionmentioning

confidence: 99%

Monitoring Potato Waste in Food Manufacturing Using Image Processing and Internet of Things Approach

et al. 2019

View full text Add to dashboard Cite

Approximately one-third of the food produced globally is spoiled or wasted in the food supply chain (FSC). Essentially, it is lost before it even reaches the end consumer. Conventional methods of food waste tracking relying on paper-based logs to collect and analyse the data are costly, laborious, and time-consuming. Hence, an automated and real-time system based on the Internet of Things (IoT) concepts is proposed to measure the overall amount of waste as well as the reasons for waste generation in real-time within the potato processing industry, by using modern image processing and load cell technologies. The images captured through a specially positioned camera are processed to identify the damaged, unusable potatoes, and a digital load cell is used to measure their weight. Subsequently, a deep learning architecture, specifically the Convolutional Neural Network (CNN), is utilised to determine a potential reason for the potato waste generation. An accuracy of 99.79% was achieved using a small set of samples during the training test. We were successful enough to achieve a training accuracy of 94.06%, a validation accuracy of 85%, and a test accuracy of 83.3% after parameter tuning. This still represents a significant improvement over manual monitoring and extraction of waste within a potato processing line. In addition, the real-time data generated by this system help actors in the production, transportation, and processing of potatoes to determine various causes of waste generation and aid in the implementation of corrective actions.

show abstract

Section: Introductionmentioning

confidence: 99%

Monitoring Potato Waste in Food Manufacturing Using Image Processing and Internet of Things Approach

et al. 2019

View full text Add to dashboard Cite

show abstract

Internet of Things: Roboboat for Water Area Monitoring using 4G network and Google Firebase

Ramadan

Hadiyoso

Sakti

2019

2019 International Conference of Artificial Intelligence and Information Technology (ICAIIT)

View full text Add to dashboard Cite

RNEP: Random Node Entropy Pairing for Efficient Decentralized Training with Non-IID Local Data

Kang,

Lee

2024

Electronics

View full text Add to dashboard Cite

The proliferation of edge devices and advancements in Internet of Things technology have created a vast array of distributed data sources, necessitating machine learning models that can operate closer to the point of data generation. Traditional centralized machine learning approaches often struggle with real-time big data applications, such as climate prediction and traffic simulation, owing to high communication costs. In this study, we introduce random node entropy pairing, a novel distributed learning method for artificial neural networks tailored to distributed computing environments. This method reduces communication overhead and addresses data imbalance by enabling nodes to exchange only the weights of their local models with randomly selected peers during each communication round, rather than sharing the entire dataset. Our findings indicate that this approach significantly reduces communication costs while maintaining accuracy, even when learning from non-IID local data. Furthermore, we explored additional learning strategies that enhance system accuracy by leveraging the characteristics of this method. The results demonstrate that random node entropy pairing is an effective and efficient solution for distributed learning in environments where communication costs and data distribution present significant challenges.

show abstract

Real Time Adaptive Approach for Image Processing Using Mobile Nodes

Cited by 3 publications

References 3 publications

Monitoring Potato Waste in Food Manufacturing Using Image Processing and Internet of Things Approach

Monitoring Potato Waste in Food Manufacturing Using Image Processing and Internet of Things Approach

Internet of Things: Roboboat for Water Area Monitoring using 4G network and Google Firebase

RNEP: Random Node Entropy Pairing for Efficient Decentralized Training with Non-IID Local Data

Contact Info

Product

Resources

About