基于改进YOLOv7的自然环境乌梅成熟度检测方法

打开文本图片集

中图分类号：S567；TP399 文献标识码：A 文章编号：2095-5553（2025）07-0124-07

Abstract：Black plumwithdiferent maturityhas differentpharmacologicalefects.Tojudgethematuritymostfruits in the orchard，a series improvementsare cariedout basedon the YOLOv7 target detection algorithm.The Vision Transformer with Bi-Level Routing Atention（BiFormer）module isadded tothe Backbone YOLOv7 modelto improve thefeatureexpresionability thenetwork.Thefruitmaturityrefinementmoduleisdesignedtoimprovethecorrectrate fruit maturity detection.The studyshows that theimprovedYOLOv7—1model hasa Mean AveragePrecision （mAP） 0.805，and is higher by4.8，12.4，0.9，0.7，12.6，1.7，5.8and12.3 percentage points， respectively，compared with the improved YOLOv7—2 model，Faster R—CNN model，YOLOv3 model，Mask R— CNN model，YOLOv5s model，YOLOv5l model，YOLOv7 model，and YOLOv8 model. The improved YOLOv7— 1 model can improve the accuracy identifying the maturity black plum.

Keywords：black plum;maturity detection;deep learning；natural environment

0 引言

成为合理安排劳动力以及适时采摘的必要条件。（剩余8694字）