Water Column Detection Method at Impact Point Based on Improved YOLOv4 Algorithm

For a long time, the water column at impact point of naval gun firing sea has mainly depended on manual detection methods for locating, which problems such as low accuracy, subjectivity and inefficiency. In order to solve above problems, this paper proposes method based an improved you-only-look-once version 4 (YOLOv4) algorithm. Firstly, detects antenna through Hoffman line constrain sensitive area in current image so improve accuracy detection; secondly, density-based spatial clustering applications with noise (DBSCAN) + K-means algorithm is used obtain better prior bounding box, input into YOLOv4 network positioning column; finally, convolutional block attention module (CBAM) added PANet structure column. The experimental results show that can effectively point.