A Comparison between YOLOV7 and YOLOV5-based detection of Combat Targets and Camouflaged Militia

Animals frequently use the camouflage method to disguise themselves. On the battlefield, it is also employed to cover military equipment. Camouflaged objects blend into their environment by adopting colors and textures that match their surroundings. Because of its size and extreme likeness to its surroundings, a camouflaged object cannot be recognized using a general detection approach, making its detection more challenging than that of a general object. This work analyzes the available approaches to address the issue and proposes a camouflage object identification algorithm based on the YOLOv7 algorithm. The self-created military camouflage target data set was used for training and testing, and compared with YOLOv5 with multiple data sets.