A Study on the Identification of Warship Type/Class by Measuring Similarity with Virtual Warship

In the defense field, accurate information about surrounding ships is important, and for this purpose, location information about nearby ships is acquired based on radio or radar. However, in a war situation, it is more important to select not only the position of the opponent, but also a high-value unit among them, and for this, an image-based warship type/class identification technology is required. This paper presents a specific method for constructing a detailed warship type/class identification system by estimating the azimuth/altitude from the input warship image, comparing the similarity using part segmentation between virtual warship model. Through the development of the proposed system, reliable warship identification information can be used for tactical applications in the defense field, contributing to strengthening defense capabilities.