Current trends on the use of deep learning methods for image analysis in energy applications
Mattia Casini, Paolo De Angelis, Eliodoro Chiavazzo, Luca Bergamasco
Abstract
Deep learning methods for image analysis are attracting increasing interest for application in a wide range of different research fields. Here we aim to systematically analyze and discuss the most relevant examples for the energy sector. To this, we perform a comprehensive literature screening on applications of deep learning methods for image analysis, classify the results in application macro-areas, and discuss the emerging trends on the available energy-related cases. The results of the analysis show that, while the exploitation of these methods for energy applications still appears to be at an early stage, the interest during the last years, in terms of number of published works, has considerably grown. To provide a systematic overview on the available energy-related examples, we present a schematic correlation chart mapping algorithms, tasks, and applications. The reported analysis is intended to provide an up-to-date overview on the current application trends and potential developments for energy applications in the next future.