부산대학교 도서관

In order to provide real-time identification of objects in Internet of Things (IoT) software for the healthcare industry, this research examines the incorporation of cutting-edge computing with convolutional neural networks (CNNs). A deductive method is used to methodically optimize CNN representations for constrained by resources edge devices, and it is based on an interpretivist mindset. To give a thorough explanation of the technical complexities concerned, a design based on descriptive statistics is used. The analysis is built on secondary data gathered from reliable sources. The topics covered range from enhancing CNN machine learning models on edge devices to putting secure processing of information practices in place, analyzing the effects on medical diagnostics particularly monitoring, and examining scalability across various healthcare contexts. Results indicate significant gains in object detection precision and decreased latency, demonstrating the viability of implementing Edge-CNN methods. Critical study underlines the possibility for live installations and the necessity for specific approaches to integration in broader healthcare organizations. Conducting clinical trials, modifying the framework for particular healthcare areas, and investigating cost-benefit assessments are all recommendations. Enhanced optimization methods, adaptive safeguards, and additional scalability investigations should all be the focus of future research. Future research around real-time analytics gives an intriguing opportunity to go beyond recognizing objects as well. Overall, this research revolutionizes real-time object identification in medical IoT for better patient care by advancing the practical use of edge technologies and CNNs.