Study on Lightweight Apple Detection Model for Smart Orchards

doi:10.15933/j.cnki.1004-3268.2026.04.014

Abstract

Abstract: In the development of automated agricultural harvesting equipment，high‑precision recognition，real‑time response，and lightweight design are core requirements for target detection algorithms.To address these technical challenges，this study innovatively proposed a lightweight apple detection model named YEMB‑FPN（YOLOv10n‑efficient multi‑branch FPN）. The model replaced the last two C2f modules in the backbone network of the original YOLOv10n with the grouped efficient multi‑scale（GEMS） convolution module. In the neck network，a bidirectional feature pyramid framework was adopted，where the original Concat operation was replaced by a weighted bidirectional feature pyramid fusion module（BiFPN_Fusion）. Additionally，a novel cross‑stage multi‑scale fusion block（CS‑MSFB）was designed to substitute all original C2f and C2fCIB modules in the neck network.The model was validated using a self‑constructed apple target detection dataset from orchard environments.Experimental results demonstrated that the YEMB‑FPN model achieved an mAP50 of 98.6%，with a model size of only 4.6 MB.Compared to the baseline YOLOv10n，this represented a 3.3 percentage point improvement in mAP50 and a reduction of 1.2 MB in model size.These findings indicated that the proposed algorithm significantly enhanced apple detection accuracy，while its lightweight design markedly improved compatibility with low‑computational hardware platforms，which provided critical technical support for the embedded deployment of intelligent agricultural equipment.

Key words: Apple, Object detection, YOLOv10n, Feature pyramid network, Multi?scale convolution, Lightweight

摘要： 在农业自动化采摘装备研发过程中，高精度识别、实时响应与轻量化设计成为目标检测算法需满足的核心指标。针对上述技术挑战，构建了轻量级苹果检测模型YEMB-FPN（YOLOv10n‑efficient multi‑branch FPN）。模型在主干网络（Backbone）使用分组高效多尺度卷积模块GEMS（Grouped efficient multi‑scale）代替原始YOLOv10n的主干网络最后2个C2f模块；在颈部网络（Neck）中，整体框架采用双向特征金字塔网络结构，使用加权双向特征金字塔融合模块（BiFPN_Fusion）替代原Neck的Concat，并创新设计跨阶段多尺度融合模块CS-MSFB（Cross‑stage multi‑scale fusion block）替代原Neck中的所有C2f和C2fCIB模块，完成模型改进。采用自建果园苹果目标检测数据集进行验证。结果表明，YEMB-FPN模型的mAP50达到98.6%，模型大小仅为4.6 MB，较基本模型YOLOv10n的mAP50提升3.3百分点，模型大小减少1.2 MB。表明该算法模型有效提升了苹果检测精度，轻量化设计显著增强了模型在低算力硬件平台的适配性。

关键词: 苹果, 目标检测, YOLOv10n, 特征金字塔网络, 多尺度卷积, 轻量化

CLC Number:

HU Junfeng, LIU Zilong, LIU Dayang. Study on Lightweight Apple Detection Model for Smart Orchards[J]. Journal of Henan Agricultural Sciences, 2026, 55(4): 140-149.

胡峻峰, 刘子龙, 刘大洋. 面向智慧果园的轻量化苹果检测模型研究[J]. 河南农业科学, 2026, 55(4): 140-149.

Figures/Tables 12

References

［1］赵琳，李星瑶，宋玉军，等. 回归分析烟台苹果影响因素［J］. 中国科技信息，2023（14）：71‑73.
ZHAO L，LI X Y，SONG Y J，et al. Regression analysis of influencing factors on Yantai apple［J］.China Science and Technology Information，2023（14）：71‑73.

［2］YAN B，FAN P，LEI X Y，et al. A real‑time apple targets detection method for picking robot based on improved YOLOv5［J］. Remote Sensing，2021，13（9）：1619.

［3］WANG C L，LI C J，HAN Q Y，et al. A performance analysis of a Litchi picking robot system for actively removing obstructions，using an artificial intelligence algorithm［J］. Agronomy，2023，13（11）：2795.

［4］WANG K P，MENG Z J，WU Z. Deep learning‑based ground target detection and tracking for aerial photography from UAVs［J］. Applied Sciences，2021，11 （18）：8434.

［5］SUN L J，HU G R，CHEN C，et al. Lightweight apple detection in complex orchards using YOLOV5‑PRE［J］.Horticulturae，2022，8（12）：1169.

［6］黄天才，刘建新，张筱晨，等.基于改进YOLOv8的采摘机器人苹果检测算法研究［J］. 数字农业与智能农机，2025（2）：14‑18.
HUANG T C，LIU J X，ZHANG X C，et al. Research on apple detection algorithm of picking robot based on improved YOLOv8［J］. Digital Agriculture and Intelligent Agricultural Machinery，2025（2）：14‑18.

［7］周晖，杨洁，赵祥飞. YOLOv8轻量化的果园复杂环境下苹果检测算法［J］. 科学技术与工程，2025，25（6）：2274‑2283.
ZHOU H， YANG J， ZHAO X F. Lightweight improvement of YOLOv8 for apple detection in complex orchard environments［J］. Science Technology and Engineering，2025，25（6）：2274‑2283.

［8］张静，张昊，聂尚卿，等. 基于改进YOLOv8的苹果采摘机器人视觉算法［J］. 液压气动与密封，2025，45（3）：36‑43.
ZHANG J，ZHANG H，NIE S Q，et al. Vision algorithm for apple picking robot based on improved YOLOv8［J］.Hydraulics Pneumatics & Seals，2025，45（3）：36‑43.

［9］WANG A，CHEN H，LIU L H，et al. YOLOv10：Real‑time end‑to‑end object detection［C］//Advances in Neural Information Processing Systems 37. Vancouver，Canada： Neural Information Processing Systems Foundation，Inc. ，2024：107984‑108011.

［10］ HUSSAIN M，KHANAM R. In‑depth review of YOLOv1 to YOLOv10 variants for enhanced photovoltaic defect detection［J］. Solar，2024，4（3）：351‑386.

［11］SUNDARESAN GEETHA A，AL RABBANI ALIF M，HUSSAIN M，et al. Comparative analysis of YOLOv8 and YOLOv10 in vehicle detection：Performance metrics and model efficacy［J］. Vehicles，2024，6（3）：1364‑1382.

［12］FU H，GUO Z W，FENG Q C，et al. MSOAR‑YOLOv10：Multi‑scale occluded apple detection for enhanced harvest robotics［J］. Horticulturae，2024，10（12）：1246.

［13］PEKER R B，KURTOGLU C O. Evaluation of the performance of a YOLOv10‑based deep learning model for tooth detection and numbering on panoramic radiographs of patients in the mixed dentition period［J］. Diagnostics，025，15（4）：405.

［14］ZHANG K X，LAMMERS K，CHU P Y，et al. An automated apple harvesting robot—From system design to field evaluation［J］. Journal of Field Robotics，2024，41（7）：2384‑2400.

［15］GAO R L，ZHOU Q J，CAO S X，et al. Apple‑picking robot picking path planning algorithm based on improved PSO［J］. Electronics，2023，12（8）：1832.

［16］HAN K，WANG Y H，XU C，et al. GhostNets on heterogeneous devices via cheap operations［J］.International Journal of Computer Vision，2022，130 （4）：1050‑1069.

［17］CHEN Y P，DAI X Y，CHEN D D，et al. Mobile‑former：Bridging MobileNet and transformer［C］//2022IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. New Orleans，LA，USA：IEEE， 2022：5260‑5269.

［18］CHEN J，MAI H S，LUO L B，et al. Effective feature fusion network in BIFPN for small object detection［C］//2021 IEEE International Conference on Image Processing（ICIP）. Anchorage，AK，USA：IEEE，2021：699‑703.

［19］YU C S，SHIN Y. SAR ship detection based on improved YOLOv5 and BiFPN［J］. ICT Express，2024，10（1）：28‑33.

［20］GULZAR Y. Fruit image classification model based on MobileNetV2 with deep transfer learning technique［J］.Sustainability，2023，15（3）：1906.

［21］ZHANG Q L，YANG Y B. SA‑net：Shuffle attention for deep convolutional neural networks［C］//ICASSP 2021‑2021 IEEE International Conference on Acoustics，Speech and Signal Processing（ICASSP）.Toronto，ON，Canada：IEEE，2021：2235‑2239.

[1]	CHEN Chen, QIAN Jinhua, ZHU Weixing, LIU Rui, JIANG Yi. Identification of Group‑Housed Pigs under the Aggression Situation Based on GBPC‑ResNeXt50 [J]. Journal of Henan Agricultural Sciences, 2026, 55(4): 150-159.
[2]	LIU Fangjie, HAN Xujuan, WANG Yuxuan. Control Efficiency and Mechanism of Exogenous Melatonin on Postharvest Apple Ring Rot [J]. Journal of Henan Agricultural Sciences, 2026, 55(1): 110-117.
[3]	NIE Lin, ZHAO Hongliang, CAO Yijing, SUN Ang, LU Hang, LIU Limin. Effects of Different Chemical Fruit Thinning Agents on Thinning and Fruit Quality of Meiba and Pink Lady Apple [J]. Journal of Henan Agricultural Sciences, 2025, 54(6): 129-134.
[4]	YANG Ximeng, ZHANG Biao, LI Jiali, CUI Xiufen, CAO Yafeng. Comprehensive Evaluation of Mid‐Early Maturing Apple Varieties Based on Grey Correlation Analysis，DTOPSIS Method，and Membership Function Method [J]. Journal of Henan Agricultural Sciences, 2025, 54(10): 121-130.
[5]	LI Xiaosuo, GUO Wang, ZHU Huaji, GU Jingqiu, LI Qingxue, WU Huarui. Quality Detection Model of Unmanned Harvesting Operation of Open‐Field Cabbage Based on Machine Vision [J]. Journal of Henan Agricultural Sciences, 2025, 54(10): 150-158.
[6]	HU Junfeng, LI Songqing, HUANG Xiaowen, LIU Dayang, LI Baicong. Lightweight Grape Object Detection Fusion Algorithm Based on Improved YOLOv5s [J]. Journal of Henan Agricultural Sciences, 2024, 53(9): 150-158.
[7]	WANG Jinxin, JIA Linguang, SHAO Jianzhu, SUN Jianshe, PENG Jianying. Changes of Carbohydrate Content and C/N in Branches and Leaves during Flower Bud Differentiation of Tianhong 2 Apple [J]. Journal of Henan Agricultural Sciences, 2024, 53(7): 124-132.
[8]	DAI Jiabing, SONG Chunfang, LING Caijin, LI Zhenfeng, SUN Chonggao. Research on Insect‑bitten Zijin Tea Detection Method Based on YOLOv5s‑SE and Channel Pruning [J]. Journal of Henan Agricultural Sciences, 2024, 53(5): 157-163.
[9]	LI Dahua, KONG Shu, LI Dong, YU Xiao. Light Weight Detection Algorithm for Apple Surface Defect Based on Improved YOLOv7 [J]. Journal of Henan Agricultural Sciences, 2024, 53(3): 141-150.
[10]	BAI Jiayun, XU Junxiang, SUN Qinping, YAO Hai, YANG Yingjie, HE Liang, GAO Shitao, XIONG Jianjun, LIU Yanpeng. Effect of Continuous Application of Municipal Standardized Sewage Sludge on the Risk of Heavy Metal Pollution in Apples and Orchard Soil [J]. Journal of Henan Agricultural Sciences, 2024, 53(11): 83-92.
[11]	CHEN Cong, YU Xiao, GONG Qi. Apple Leaf Diseases Identification Based on Improved Residual Network [J]. Journal of Henan Agricultural Sciences, 2023, 52(4): 152-161.
[12]	ZHANG Sipu, ZHOU Qianqian, CUI Wei, ZHANG Ke, HAN Lixin, QU Zhenfang, LU Yunfeng, TANG Cunduo, NIU Jiajia. Effects of L‐tert‐leucine on Quality and Storage of Gala Apple [J]. Journal of Henan Agricultural Sciences, 2023, 52(12): 172-180.
[13]	WANG Yingchao, ZHANG Jingjing, JIA Donglin, ZHOU Tengfei. Grading of Apple Based on K‑means Clustering and Improved MLP [J]. Journal of Henan Agricultural Sciences, 2023, 52(1): 161-171.
[14]	YANG Yang, LIU Yuanyuan. Identification and Sequence Analysis of the MdLAC Gene Family Members in Apple [J]. Journal of Henan Agricultural Sciences, 2021, 50(7): 125-135.
[15]	NIU Jiajia, ZHANG Sipu, ZHANG Ke, HAN Lixin, QU Zhenfang, MIAO Jianyin. Effects of 5‑Aminolevulinic Acid Treatment on Storage of Fuji Apple during Fruit Expansion Stage [J]. Journal of Henan Agricultural Sciences, 2021, 50(10): 146-153.