Military Image Scene Recognition Based on CNN and Semantic Information

Abstract

Due to the complexity of the military scene and resulted low accuracy in military scene recognition, this paper proposes a military image scene recognition method based on CNN and semantic information. First, the scene image is initially classified by CNN, and then the classification result is optimized by using the semantic relationship between the military target and the scene. This paper uses the collected military image scene recognition dataset to evaluate the proposed method. The experimental results show that the proposed method has better accuracy than the unmodified CNN algorithm and achieve good recognition results on the test set, which has potential for future application.