AUTOMATIC DETECTION AND CLASSIFICATION OF AIRCRAFT IN IMAGES: A CASE STUDY USING DEEP LEARNING TECHNIQUES

The detection and classification of targets in images are part of the planning stages for an operation within the context of Intelligence, providing situational awareness to decision-makers. In Surveillance and Reconnaissance actions, target detection can be performed using Deep Neural Networks, making the process automatic and reliable. This work proposes to detect and identify aircraft in images in the visible spectrum using the YOLOv7 network, a deep learning technique, aimed at minimizing automation issues with large volumes of data. The methodology employs appropriate pre-training of the network using a previously assembled dataset. Finally, metrics are used to evaluate the results, demonstrating the potential to serve as a methodological guide for future studies and application in the automatic detection of other types of targets or features in real time.