Improvement of ship target detection algorithm for YOLOv7‐tiny

• Published in: IET image processing Vol. 18; no. 7; pp. 1710 - 1718
• Main Authors: Zhang, Huixia, Yu, Haishen, Tao, Yadong, Zhu, Wenliang, Zhang, Kaige
• Format: Journal Article
• Language: English
• Published: Wiley 01.05.2024
• Subjects: image classification; image recognition
• ISSN: 1751-9659; 1751-9667; 1751-9667
• DOI: 10.1049/ipr2.13054

In addressing the challenge of ships being prone to occlusion in multi‐target situations during ship target detection, leading to missed and false detections, this paper proposes an enhanced ship detection algorithm for YOLOv7‐tiny. The proposed method incorporates several key modifications. Firstly, it introduces the Convolutional Block Attention Module in the Backbone section of the original model, emphasizing position information while attending to channel features to enhance the network's ability to extract crucial information. Secondly, it replaces standard convolution with GSConv convolution in the Neck section, preserving detailed information and reducing computational load. Subsequently, the lightweight operator Content‐Aware ReAssembly of Features is employed to replace the original nearest‐neighbour interpolation, mitigating the loss of feature information during the up‐sampling process. Finally, the localization loss function, SIOU Loss, is utilized to calculate loss, expedite training convergence, and enhance detection accuracy. The research results indicate that the precision of the improved model is 91.2%, mAP@0.5 is 94.5%, and the F1‐score is 90.7%. These values are 3.7%, 5.5%, and 4.2% higher than those of the original YOLOv7‐tiny model, respectively. The improved model effectively enhances detection accuracy. Additionally, the improved model achieves an FPS of 145.4, meeting real‐time requirements. This paper presents an enhanced ship detection algorithm for YOLOv7‐tiny, addressing occlusion challenges in multi‐target ship detection. The modifications include integrating Convolutional Block Attention Module for position emphasis, GSConv in the Neck section, Content‐Aware ReAssembly of Features for up‐sampling, and SIOU Loss for better accuracy. The improved model achieves a precision of 91.2%, mAP@0.5 of 94.5%, and an F1‐score of 90.7%, surpassing the original model by 3.7%, 5.5%, and 4.2%, respectively, significantly boosting detection accuracy.