YOLOv11 and Faster R-CNN for Palm Fruit Ripeness Detection: Consistent Class-Level Difficulty with Explainable AI Insights

Queenta Paradissa Ramadhani; Favian Dewanta; Teddy Surya Gunawan

doi:10.15408/jti.v19i2.53853

Authors

Queenta Paradissa Ramadhani Department of Electrical Engineering, Telkom University, Indonesia
Favian Dewanta Department of Electrical Engineering, Telkom University, Indonesia
Teddy Surya Gunawan Department of Electrical and Computer Engineering, International Islamic University Malaysia, Malaysia

DOI:

https://doi.org/10.15408/jti.v19i2.53853

Keywords:

YOLOv11, Faster R-CNN, Deep Learning, Object Detection, Explainable AI, Precision Agriculture

Abstract

Direct comparison between one-stage and two-stage object detectors under matched training conditions remains limited for palm oil Fresh Fruit Bunch (FFB) ripeness detection. This study compares five YOLOv11 variants and two Faster R-CNN variants using 16,842 annotated instances from Roboflow Universe across six ripeness categories under a controlled training configuration. YOLOv11m achieved the best overall performance with an mAP50-95 of 89.00%, outperforming Faster R-CNN R-101-FPN (85.91%) by 3.09 percentage points, while providing higher inference speed (64.42 versus 23.30 FPS) and a substantially smaller model size (38.65 versus 459.93 MB). Repeated evaluation across three random seeds showed stable performance, with mean mAP50-95 values of 89.01 ± 0.19% for YOLOv11m and 86.12 ± 0.38% for R-101-FPN. Both representative models exhibited the same class-level difficulty pattern, with Unripe achieving the highest AP and Abnormal the lowest. XGrad-CAM analysis further showed more dispersed activation patterns for the more difficult categories. These findings indicate that YOLOv11m provides the most favorable accuracy-speed-model-size trade-off among the evaluated models, while improving difficult categories may also require targeted data collection and annotation refinement.

References

[1] H. Varkkey, A. Tyson, and S. A. B. Choiruzzad, “Palm oil intensification and expansion in Indonesia and Malaysia: Environmental and socio-political factors influencing policy,” For. Policy Econ., vol. 92, pp. 148–159, Jul. 2018, doi: 10.1016/j.forpol.2018.05.002.

[2] N. D. Hadi, B. L. Handoko, A. S. L. Lindawati, and H. Sarjono, “Palm Oil Industry Dynamics: Assessing P/B Ratios of Indonesian Palm Oil Companies through Palm Tree Profile, Average FFB Yield, and Palm Oil Extraction Rate,” Journal of Applied Data Sciences, vol. 6, no. 1, pp. 236–246, Jan. 2025, doi: 10.47738/jads.v6i1.483.

[3] J. W. Lai, H. R. Ramli, L. I. Ismail, and W. Z. Wan Hasan, “Oil Palm Fresh Fruit Bunch Ripeness Detection Methods: A Systematic Review,” Agriculture (Switzerland), vol. 13, no. 1, pp. 156–172, Jan. 2023, doi: 10.3390/agriculture13010156.

[4] N. Paul, G. C. Sunil, D. Horvath, and X. Sun, “Deep learning for plant stress detection: A comprehensive review of technologies, challenges, and future directions,” Comput. Electron. Agric., vol. 229, no. 1, pp. 1–23, Feb. 2025, doi: 10.1016/j.compag.2024.109734.

[5] N. Mohamad Hamidi, A. R. Saili, and S. Syahlan, “A Systematic Literature Review of Factors Affecting the Yield of Fresh Fruit Bunch and Oil Extraction Rate among Smallholders in Malaysia,” International Journal of Academic Research in Business and Social Sciences, vol. 14, no. 9, pp. 214–229, Sep. 2024, doi: 10.6007/ijarbss/v14-i9/22274.

[6] A. P. G. Kadir et al., “Oil Palm Economic Performance in Malaysia and R&D Progress in 2022,” J. Oil Palm Res., vol. 35, no. 2, pp. 193–216, Jun. 2023, doi: 10.21894/jopr.2023.0028.

[7] T. S. Gunawan, M. Kartiwi, H. Mansor, and N. M. Yusoff, “Palm Fruit Ripeness Detection and Classification Using Various YOLOv8 Models,” in ICSIMA 2023 - 9th IEEE International Conference on Smart Instrumentation, Measurement and Applications, Kuala Lumpur, Malaysia: Institute of Electrical and Electronics Engineers Inc., Nov. 2023, pp. 193–198. doi: 10.1109/ICSIMA59853.2023.10373435.

[8] H. Ma et al., “YOLOv11-GSF: an optimized deep learning model for strawberry ripeness detection in agriculture,” Front. Plant Sci., vol. 16, 2025, doi: 10.3389/fpls.2025.1584669.

[9] P. Kamat, S. Gite, H. Chandekar, L. Dlima, and B. Pradhan, “Multi-class fruit ripeness detection using YOLO and SSD object detection models,” Discover Applied Sciences, vol. 7, no. 9, Sep. 2025, doi: 10.1007/s42452-025-07617-7.

[10] J. Josdaan, V. C. Tamsil, J. Harefa, and K. Jingga, “Revolutionizing palm oil ripeness classification: Utilizing YOLOv8 for ultra-precise ripeness detection,” in Procedia Computer Science, Elsevier B.V., 2024, pp. 700–709. doi: 10.1016/j.procs.2024.10.296.

[11] J. Y. Goh, Y. Md Yunos, and M. S. Mohamed Ali, “Fresh Fruit Bunch Ripeness Classification Methods: A Review,” Food and Bioprocess Technology, vol. 18, no. 1, pp. 183–206, Jun. 2024, doi: 10.1007/s11947-024-03483-0.

[12] E. Edozie, A. N. Shuaibu, U. K. John, and B. O. Sadiq, “Comprehensive review of recent developments in visual object detection based on deep learning,” Artif. Intell. Rev., vol. 58, no. 9, pp. 1–97, Sep. 2025, doi: 10.1007/s10462-025-11284-w.

[13] S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 39, no. 6, pp. 1137–1149, Jun. 2017, doi: 10.1109/TPAMI.2016.2577031.

[14] A. A. J. Pazhani and C. Vasanthanayaki, “Object detection in satellite images by faster R-CNN incorporated with enhanced ROI pooling (FrRNet-ERoI) framework,” Earth Sci. Inform., vol. 15, no. 1, pp. 553–561, Mar. 2022, doi: 10.1007/s12145-021-00746-8.

[15] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA: IEEE, Jun. 2016, pp. 779–788. doi: 10.1109/CVPR.2016.91.

[16] M. L. Ali and Z. Zhang, "The YOLO Framework: A Comprehensive Review of Evolution, Applications, and Benchmarks in Object Detection," Computers, vol. 13, no. 12, p. 336, Dec. 2024, doi: 10.3390/computers13120336.

[17] J. Terven, D. M. Córdova-Esparza, and J. A. Romero-González, “A Comprehensive Review of YOLO Architectures in Computer Vision: From YOLOv1 to YOLOv8 and YOLO-NAS,” Mach. Learn. Knowl. Extr., vol. 5, no. 4, pp. 1680–1716, Dec. 2023, doi: 10.3390/make5040083.

[18] S. Gholinavaz, N. saeedi, and S. S. Gharehveran, “Robustness analysis of YOLO and Faster R-CNN for Object detection in Realistic Weather Scenarios with Noise Augmentation,” Sci. Rep., vol. 15, no. 1, pp. 1–27, Dec. 2025, doi: 10.1038/s41598-025-28737-5.

[19] G. Jocher and R. Munawar, “Ultralytics YOLO11,” Ultralytics Inc. Accessed: Jun. 27, 2026. [Online]. Available: https://docs.ultralytics.com/models/yolo11#overview

[20] M. Dalal and P. Mittal, "A Systematic Review of Deep Learning-Based Object Detection in Agriculture: Methods, Challenges, and Future Directions," Computers, Materials and Continua, vol. 84, no. 1, pp. 57–91, Jun. 2025, doi: 10.32604/cmc.2025.066056.

[21] S. A. Ghazalli, H. Selamat, N. Khamis, and M. F. Haniff, “Short Review on Palm Oil Fresh Fruit Bunches Ripeness and Classification Technique,” Journal of Advanced Research in Applied Mechanics, vol. 106, no. 1, pp. 37–47, 2023, doi: 10.37934/aram.106.1.3747.

[22] T. Shehzadi, K. A. Hashmi, M. Liwicki, D. Stricker, and M. Z. Afzal, “Object Detection with Transformers: A Review,” Sensors, vol. 25, no. 19, p. 6025, Oct. 2025, doi: 10.3390/s25196025.

[23] G. Liu, D. Yang, J. Ye, H. Lu, Z. Wang, and Y. Zhao, “A real-time welding defect detection framework based on RT-DETR deep neural network,” Advanced Engineering Informatics, vol. 65, p. 103318, May 2025, doi: 10.1016/j.aei.2025.103318.

[24] Q. Paradissa, “Palm Fruit Ripeness Dataset,” Roboflow Universe. Accessed: Mar. 28, 2026. [Online]. Available: https://universe.roboflow.com/queenta-paradissa/dataset-6-4260-citra

[25] J. Bento, T. Paixão, and A. B. Alvarez, "Performance Evaluation of YOLOv8, YOLOv9, YOLOv10, and YOLOv11 for Stamp Detection in Scanned Documents," Applied Sciences, vol. 15, no. 6, p. 3154, Mar. 2025, doi: 10.3390/app15063154.

[26] K. He, G. Gkioxari, P. Dollar, and R. Girshick, “Mask R-CNN,” in 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy: IEEE, Oct. 2017, pp. 2980–2988. doi: 10.1109/ICCV.2017.322.

[27] T.-Y. Lin, P. Dollar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature Pyramid Networks for Object Detection,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA: IEEE, Jul. 2017, pp. 936–944. doi: 10.1109/CVPR.2017.106.

[28] R. Padilla, S. L. Netto, and E. A. B. da Silva, “A Survey on Performance Metrics for Object-Detection Algorithms,” in 2020 International Conference on Systems, Signals and Image Processing (IWSSIP), Brazil: IEEE, Jul. 2020, pp. 237–242. doi: 10.1109/IWSSIP48289.2020.9145130.

[29] G. Lavanya and S. D. Pande, “Enhancing Real-time Object Detection with YOLO Algorithm,” Journal of EAI Endorsed Transactions on Internet of Things, vol. 10, no. 1, pp. 1–9, Dec. 2023, doi: 10.4108/eetiot.4541.

[30] H. Zhang and X. Yang, “A Survey on Algorithms and Performance Metrics for Object Detection,” International Core Journal of Engineering, vol. 9, no. 9, pp. 113–145, Sep. 2023, doi: 10.6919/ICJE.202309_9(9).0017.

[31] R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, "Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization," in 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy: IEEE, Oct. 2017, pp. 618–626. doi: 10.1109/ICCV.2017.74.

[32] Z. Feng, M. Zhu, L. Stanković, and H. Ji, “Self-Matching CAM: A Novel Accurate Visual Explanation of CNNs for SAR Image Interpretation,” Remote Sensing, vol. 13, no. 9, p. 1772, May 2021, doi: 10.3390/rs13091772.

[33] R. Fu, Q. Hu, X. Dong, Y. Guo, Y. Gao, and B. Li, “Axiom-based Grad-CAM: Towards Accurate Visualization and Explanation of CNNs,” in Proceedings of the British Machine Vision Conference 2020, Manchester, England: British Machine Vision Association, Jan. 2020. doi: 10.5244/C.34.146.

[34] X. A. Inbaraj, C. Villavicencio, J. J. Macrohon, J.-H. Jeng, and J.-G. Hsieh, "Object Identification and Localization Using Grad-CAM++ with Mask Regional Convolution Neural Network," Electronics, vol. 10, no. 13, p. 1541, Jun. 2021, doi: 10.3390/electronics10131541.