基于CDD-YOLO的轻量级低光照目标检测算法  

Lightweight Low-Light Object Detection Algorithm Based on CDD-YOLO

作　　者：史丽晨[1] 杨超 刘雪超[1] 周星宇 SHI Lichen;YANG Chao;LIU Xuechao;ZHOU Xingyu(School of Mechanical and Electrical Engineering,Xi’an University of Architecture and Technology,Xi’an 710055,China)

机构地区：[1]西安建筑科技大学机电工程学院,西安710055

出　　处：《计算机工程与应用》2025年第6期106-117,共12页Computer Engineering and Applications

基　　金：陕西省重点研发项目(2023-YBGY-386)。

摘　　要：针对低照度场景下目标检测算法面临的检测精度不高、计算成本以及内存消耗大等问题,提出一种改进YOLOv8的轻量级低光照目标检测网络模型CDD-YOLO。提出一个基于坐标注意力机制的多尺度卷积模块,提取不同感受野纹理特征并捕获空间位置之间的远程依赖关系;将动态头部框架集成到检测头中,减少复杂背景和尺度变化的干扰;基于动态非单调聚焦机制设计边界框回归损失函数,提升锚框回归路径和质量,提高模型对光照变化和噪声的适应能力;通过剪枝算法修剪模型中的冗余参数,实现模型轻量化。采用自建数据集、ExDark和VOC数据集进行实验验证,实验结果表明该方法与主流算法相比具有更好的检测效果,在计算复杂度与检测精度之间实现了更好的平衡。To address the challenges of low detection accuracy,high computational costs,and excessive memory consumption encountered by target detection algorithms in low-light conditions,this paper proposes a lightweight low-light target detection network model,CDD-YOLO,to enhance the performance of YOLOv8.Firstly,a multi-scale convolutional module based on a coordinate attention mechanism is proposed to extract texture features from different sensory fields and to capture long-range dependencies between spatial locations.Secondly,a dynamic head frame is integrated into the detection head to minimize the interference caused by complex backgrounds and scale variations.The bounding box regression loss function is designed using a dynamic non-monotonic focusing mechanism to enhance the regression path and quality of the anchor boxes,thereby improving the adaptability of model to variations in lighting and noise.Finally,redundant parameters in the model are pruned using a pruning algorithm to achieve model lightweighting.The selfconstructed dataset,ExDark,and the VOC dataset are used for experimental validation.The experimental results show that the proposed method has better detection effect compared with the mainstream algorithms,and achieves a better balance between computational complexity and detection accuracy.

关 键 词：低照度 YOLOv8 注意力机制 损失函数 轻量化网络 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]