引用本文:
【打印本页】   【下载PDF全文】   查看/发表评论  下载PDF阅读器  关闭
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 209次   下载 186 本文二维码信息
码上扫一扫!
分享到: 微信 更多
基于改进 YOLOv7 的猕猴桃果实识别优化
何翔, 朱洪前
中南林业科技大学
摘要:
为了解决猕猴桃 Actinidia chinensis 果实识别过程中存在果实之间重叠导致的遮挡严重、 检测结 果易受叶片影响等问题, 建立不同日照条件下的猕猴桃果实图像数据集, 对 YOLOv7 模型做了 3 方面改 进: 将 Backbone 部分的卷积模块替换成 GhostConv 模块, 在维持原有精度的程度上降低模型的参数量; 针对猕猴桃果实之间存在大量重叠的情况, 引入非极大值抑制 NMS (Soft-NMS) 策略提高检测框回归精 度; 融合 SimAM 注意力机制, 增强模型对于高密度猕猴桃特征的提取能力。 通过对比实验表明, 优化后 的模型与 Faster RCNN 相比, mAP 值增加了 12. 7 个百分点, 检测速度提升 106. 8 帧/ s, 综合性能较好, 满足机器实时对于猕猴桃果实识别的需求。
关键词:  果实识别  YOLOv7  GhostConv  Soft NMS  SimAM
DOI:
分类号:
基金项目:
Optimization of Actinidia chinensis Fruit Recognition based on Improved YOLOv7
hexiang, zhu hongqian
中南林业科技大学
Abstract:
In order to solve the problems of severe occlusion caused by overlapping fruits and susceptibility to leaf influence in the recognition process of Actinidia chinensis fruit, A. chinensis fruit image dataset was established under different sunlight conditions. Three improvements were made to the YOLOv7 model: replacing the convolutional module of the Backbone part with the GhostConv module, reducing the number of model parameters while maintaining the original accuracy; to address the significant overlap between A. chinensis fruits, a Non Maximum Suppression NMS (Soft NMS) strategy is introduced to improve the accuracy of detection box regression; integrating SimAM attention mechanism to enhance the model??s ability to extract high-density A. chinensis fruit features. Through comparative experiments, it was shown that the optimized model increased mAP value by 12. 7% and detection speed by 106. 8 frames/ s compared to Faster RCNN. The overall performance is good and meets the real-time recognition needs of machines for A. chinensis fruit.
Key words:  fruit recognition  YOLOv7  GhostConv  Soft NMS  SimAM