基于改进的YOLOv8草莓成熟度检测算法研究

Research on Strawberry Maturity Detection Algorithm Based on Improved YOLOv8

摘要: 为克服传统草莓成熟度识别方式的效率与准确性局限，本研究开发了一种高效、精准且结构轻量的自动化检测模型，以促进智慧农业在草莓生产中的应用。本研究设计了一种基于改进YOLOv8框架的轻量化识别系统。该系统以YOLOv8s为基础，在主干网络末端引入通道–空间双路注意力机制(CBAM)，通过自适应权重分配增强对草莓果面光泽和种子色泽等关键性状的感知能力。同时，将原C2f模块替换为FasterNet中的C2f_Faster轻量型结构，从而在维持特征提取性能的前提下精简模型参数。优化模型在草莓成熟度判别任务中表现优异。最终检测精度达97.2%，较原YOLOv8s模型提升0.8个百分点。模型参数量减少16.2%，计算量降低32.7%，实现了轻量化目标。在遮挡场景下，小目标检测精度提升0.3个百分点。该方法可快速准确检测温室环境下的草莓果实成熟度，在精度和效率上均优于原模型，为草莓种植中的智慧农业应用提供有效技术解决方案。

Abstract: To address the limitations of conventional strawberry maturity assessment methods in terms of efficiency and accuracy, this study develops an automated detection model that is efficient, precise, and structurally lightweight, thereby facilitating the application of smart agriculture in strawberry production. A lightweight recognition system based on an improved YOLOv8 framework was designed. Using YOLOv8s as the base architecture, a Channel and Spatial Dual-Path Attention Mechanism (CBAM) was introduced at the end of the backbone network, enhancing perception of key traits such as fruit surface gloss and seed coloration through adaptive weight allocation. Simultaneously, the original C2f module was replaced with the C2f_Faster lightweight structure from FasterNet, thereby reducing model parameters while maintaining feature extraction performance. The optimized model performed excellently in strawberry maturity discrimination tasks. The final detection accuracy reached 97.2%, an improvement of 0.8 percentage points over the original YOLOv8s model. The parameter count was reduced by 16.2%, and computational load decreased by 32.7%, achieving lightweight objectives. In occluded scenarios, small target detection accuracy improved by 0.3 percentage points. This method enables rapid and accurate detection of strawberry fruit maturity in greenhouse environments, outperforming the original model in both accuracy and efficiency, and provides an effective technical solution for smart agriculture applications in strawberry cultivation.