Bowen Zheng, Huacai Lu, Shengbo Zhu, Xinqiang Chen, Hongwei Xing
Key Laboratory of Electric Drive and Control of Anhui Province, AnHui Polytechnic University, Wuhu, China.
Sci Rep. 2024 Jul 9;14(1):15771. doi: 10.1038/s41598-024-66842-z.
Aiming at the problems of error detection and missing detection in night target detection, this paper proposes a night target detection algorithm based on YOLOv7(You Only Look Once v7). The algorithm proposed in this paper preprocesses images by means of square equalization and Gamma transform. The GSConv(Group Separable Convolution) module is introduced to reduce the number of parameters and the amount of calculation to improve the detection effect. ShuffleNetv2_×1.5 is introduced as the feature extraction Network to reduce the number of Network parameters while maintaining high tracking accuracy. The hard-swish activation function is adopted to greatly reduce the delay cost. At last, Scylla Intersection over Union function is used instead of Efficient Intersection over Union function to optimize the loss function and improve the robustness. Experimental results demonstrate that the average detection accuracy of the proposed improved YOLOv7 model is 88.1%. It can effectively improve the detection accuracy and accuracy of night target detection.
针对夜间目标检测中存在的误检和漏检问题，本文提出了一种基于YOLOv7（你只看一次v7）的夜间目标检测算法。本文提出的算法通过平方均衡化和伽马变换对图像进行预处理。引入GSConv（分组可分离卷积）模块以减少参数数量和计算量，从而提高检测效果。引入ShuffleNetv2_×1.5作为特征提取网络，在保持高跟踪精度的同时减少网络参数数量。采用硬激活函数大大降低延迟成本。最后，使用Scylla交并比函数代替Efficient交并比函数来优化损失函数并提高鲁棒性。实验结果表明，所提出的改进YOLOv7模型的平均检测准确率为88.1%。它可以有效提高夜间目标检测的准确率和精度。
