In this paper, we consider the differential equation f''+ Af'+ Bf = 0, where A(z) and B(z) ≡ 0are entire functions. Assume that A(z) has a finite deficient value, then we will give some conditions on B(z)...In this paper, we consider the differential equation f''+ Af'+ Bf = 0, where A(z) and B(z) ≡ 0are entire functions. Assume that A(z) has a finite deficient value, then we will give some conditions on B(z)which can guarantee that every solution f ≡ 0 of the equation has infinite order.展开更多
针对传统算法依赖于对红外目标与环境背景的精确分离和信息提取,难以满足复杂背景和噪声等干扰因素下的检测需求。论文提出一种基于改进YOLOv5(You Only Look Once)的复杂背景红外弱小目标检测算法。该算法在YOLOv5基础上,添加注意力机...针对传统算法依赖于对红外目标与环境背景的精确分离和信息提取,难以满足复杂背景和噪声等干扰因素下的检测需求。论文提出一种基于改进YOLOv5(You Only Look Once)的复杂背景红外弱小目标检测算法。该算法在YOLOv5基础上,添加注意力机制提高算法的特征提取能力和检测效率,同时改进原YOLOv5目标检测网络的损失函数和预测框的筛选方式提高算法对红外弱小目标检测的准确率。实验选取了来自不同复杂背景的7组红外弱小目标数据集,将这些图像数据集进行标注并训练,得到红外弱小目标检测模型,然后从模型训练结果和目标检测结果的角度评估算法和模型的正确性。实验结果表明:改进的YOLOv5算法训练出来的模型,检测准确性和检测速度对比实验列出的几种目标检测算法均有明显的提升,平均精度均值(mean Average Precision,m AP)可达99.6%以上,在不同复杂背景下均可有效检测出红外弱小目标,且漏警率、虚警率低。展开更多
文摘In this paper, we consider the differential equation f''+ Af'+ Bf = 0, where A(z) and B(z) ≡ 0are entire functions. Assume that A(z) has a finite deficient value, then we will give some conditions on B(z)which can guarantee that every solution f ≡ 0 of the equation has infinite order.
文摘针对传统算法依赖于对红外目标与环境背景的精确分离和信息提取,难以满足复杂背景和噪声等干扰因素下的检测需求。论文提出一种基于改进YOLOv5(You Only Look Once)的复杂背景红外弱小目标检测算法。该算法在YOLOv5基础上,添加注意力机制提高算法的特征提取能力和检测效率,同时改进原YOLOv5目标检测网络的损失函数和预测框的筛选方式提高算法对红外弱小目标检测的准确率。实验选取了来自不同复杂背景的7组红外弱小目标数据集,将这些图像数据集进行标注并训练,得到红外弱小目标检测模型,然后从模型训练结果和目标检测结果的角度评估算法和模型的正确性。实验结果表明:改进的YOLOv5算法训练出来的模型,检测准确性和检测速度对比实验列出的几种目标检测算法均有明显的提升,平均精度均值(mean Average Precision,m AP)可达99.6%以上,在不同复杂背景下均可有效检测出红外弱小目标,且漏警率、虚警率低。
