随着智能交通系统的快速发展,交通标志检测在自动驾驶和辅助驾驶系统中扮演着至关重要的角色。为了应对交通标志检测对低延时和高精确度的要求,本文在YOLOv8n模型的基础上进行了改进,旨在提升检测速度和准确率。首先,从提升检测速度的角度出发,本文提出了一种轻量化的卷积模块GhostConv,用于替代YOLOv8模型中的原始Conv模块,同时引入GhostC2f结构替换原始的C2f结构,以进一步减少计算复杂度并加速推理过程。其次,从提升检测准确率的角度出发,本文提出设计了SPPF-LSKA模块,增强了模型对交通标志特征的提取能力。此外,本文还提出使用SIou损失函数替换原始的Ciou损失函数,以更好地优化边界框回归,进一步提升检测精度。实验结果表明,改进后的模型在保持较低延时的同时,显著提升了交通标志检测的准确率,能够更好地满足实际应用需求。With the rapid development of intelligent transportation systems, traffic sign detection plays a critical role in autonomous and assisted driving systems. To address the requirements of low latency and high accuracy in traffic sign detection, this paper improves the YOLOv8n model by focusing on enhancing detection speed and accuracy. First, from the perspective of improving detection speed, we propose a lightweight GhostConv module to replace the original Conv module in YOLOv8, and introduce the GhostC2f structure to substitute the original C2f structure, thereby reducing computational complexity and accelerating inference. Second, in this paper, the sppf-lska module is designed to enhance the ability of the model to extract the characteristics of traffic signs. Additionally, we employ the SIoU loss function instead of the original CIoU loss function to optimize bounding box regression and further improve detection precision. Experimental results demonstrate that the improved model achieves significantly higher accuracy while maintaining low latency, better
为了提高蜣螂优化算法(DBO)全局搜索过程中的种群多样性和避免陷入局部最优的风险,利用混沌理论和精英反向学习策略提出了一种基于精英反向学习策略的改进蜣螂优化算法(EoDBO)。首先,在蜣螂初始化种群个体位置时引入Sinusoidal map混沌映射策略,以提高寻优前蜣螂种群整体质量,利于加快全局搜索速度;其次,在算法后期采用精英反向学习策略,对部分较优的蜣螂位置进行扰动以调高算法的局部开发能力。利用12个国际基准测试函数测试改进算法的性能,并于DBO算法、麻雀搜索算法(SSA)、灰狼优化算法(GWO)、鲸鱼优化算法(WOA)、黑猩猩优化算法(ChOA)进行对比分析,实验表明EoDBO算法在收敛精度和算法的稳定性方面均变现更优,且收敛速度更快。In order to improve the population diversity and avoid the risk of falling into local optimum during the global search of Dung Beetle Optimizer (DBO), an improved Dung Beetle Optimizer based on Elite Opposition Learning Strategy (EoDBO) is proposed using chaos theory and elite opposition learning strategy. First, a Sinusoidal map chaotic mapping strategy is introduced in the initialization of individual population positions of dung beetles to improve the overall quality of the dung beetle population before the search for optimality and to facilitate faster global search;second, an elite opposition learning strategy is used in the later stage of the algorithm to perturb some of the better dung beetle positions to tune up the local exploitation capability of the algorithm. The performance of the improved algorithm is tested using 12 international benchmarking functions and compared with DBO algorithm, Sparrow Search Algorithm (SSA), Gray Wolf Optimization Algorithm (GWO), Whale Optimization Algorithm (WOA), and Chimpanzee Optimization Algorithm (ChOA) for analysis. The experiments show that the EoDBO algorithm turns out to be superior in terms of convergence accuracy and stability of the al
