Comparing and Selecting The Performance Efficiency Of Computing Techniques For Real-Time Image Processing

Abstract

Recently, increase in the importance of real-time data processing, it is essential to choose an efficient computing method. This study finds the optimal computing method by comparing the performance of cloud computing (CC), edge computing (EC), and edge computing (ODEC) using on-device edge computing (ODEC). In particular, we compare the strengths and weaknesses of each technology by taking examples in areas where fast data processing and real-time response are important, such as real-time video processing. According to the research results, it can be seen that cloud computing greatly increases latency due to bottlenecks in data transmission, while edge computing and on-device AI technology can minimize latency thanks to distributed structures. It compares the performance of each technology at various data scales and emphasizes that on-device AI-based approaches perform well in environments that are less affected by the network and require large-capacity data processing and real-time response. This presents the possibility of overcoming the limitations of existing computing models and developing into smarter systems.