Minseek Kim

Digital phenotyping is increasingly recognized as an essential tool for the quantitative analysis of fungal morphology, particularly in controlled indoor cultivation systems where large numbers of fruiting bodies must be assessed consistently and non-destructively. While YOLOv8-based deep learning approaches have previously been applied in phenotypic analyses of edible mushrooms, the applicability of newer YOLO architectures to fungal phenotyping remains largely unexplored. In this study, we present a controlled-environment digital phenotyping framework for indoor mushroom cultivation and conduct a systematic benchmarking evaluation of YOLOv11 for phenotypic segmentation in comparison with YOLOv8. Using bottle-cultivated Pleurotus ostreatus and Flammulina velutipes as representative edible basidiomycetes, we performed a controlled comparison of YOLOv8-seg and YOLOv11-seg using identical datasets, preprocessing pipelines, and hyperparameter configurations. The results demonstrate that YOLOv11 achieves segmentation performance comparable to that of YOLOv8 across all evaluated metrics (ΔmAP_50–95 < 0.01) while substantially reducing computational complexity, including fewer trainable parameters, lower FLOPs, and decreased gradient load. Validation against caliper-based physical measurements revealed moderate, trait-dependent agreement, whereas inter-model consistency between YOLOv8 and YOLOv11 remained consistently high across diverse morphological and segmentation scenarios. These findings suggest that recent developments in object detection architectures can improve computational efficiency without compromising phenotypic measurement fidelity. More broadly, this study highlights the importance of periodically evaluating emerging detection architectures within biological phenotyping pipelines to ensure scalable, sustainable, and high-throughput fungal phenotyping under controlled-environment cultivation systems.

Digital phenotyping is increasingly recognized as a critical tool for quantitative analysis of fungal morphology, particularly in controlled indoor cultivation systems where large numbers of fruiting bodies must be assessed consistently and non-destructively. While YOLOv8-based deep learning approaches have been previously applied to phenotypic analysis of edible mushrooms, the applicability of newer YOLO architectures to fungal phenotyping remains unexplored. In this study, we present a controlled-environment digital phenotyping framework for indoor mushroom cultivation and evaluate the feasibility of YOLOv11 for phenotypic analysis through a direct comparison with YOLOv8. Using bottle-cultivated Pleurotus ostreatus and Flammulina velutipes as representative edible basidiomycetes, we conducted a systematic comparison of YOLOv8-seg and YOLOv11-seg under identical datasets, preprocessing pipelines, and hyperparameter configurations. The results demonstrate that YOLOv11 achieves segmentation performance comparable to YOLOv8 across all evaluated metrics (ΔmAP50–95 < 0.01), while substantially reducing computational complexity, including fewer trainable parameters, lower FLOPs, and decreased gradient load. Validation against caliper-based physical measurements revealed moderate, trait-dependent agreement, whereas inter-model consistency between YOLOv8 and YOLOv11 was consistently high across diverse morphological and segmentation scenarios. These findings indicate that architectural refinements in YOLOv11 primarily enhance computational efficiency and training behavior without altering phenotypic interpretation. Collectively, this study provides the first validation of YOLOv11 for mushroom phenotyping and highlights its potential as an efficient analytical tool for future high-throughput fungal morphology studies.

Mushrooms have long been economically and nutritionally important crops, and recent advances in digital agriculture have increased interest in automating phenotypic evaluation. Due to the limitation of traditional phenotype assessment, various artificial intelligence (AI) models including YOLOv8 have been introduced to evaluate mushroom phenotypes non-destructively and efficiently. However, unlike previous models, few studies of mushroom phenotype assessment with YOLOv11 were published. In this study, using Pleurotus ostreatus and Flammulina velutipes, comparison of mushroom phenotype analysis between YOLOv8 and YOLOv11 was processed. All images were captured under controlled conditions and conducted to be preprocessed for the model evaluation. The results demonstrated that YOLOv11 achieved segmentation accuracy comparable to YOLOv8 (ΔmAP50–95 < 0.01) while substantially improving computational efficiency with a reduction of approximately 15–20%. In validation with the physical measurements of mushroom phenotype, both models showed biologically meaningful and moderate correlations across phenotypic traits (r ≈ 0.2–0.44; R² ≈ 0.72–0.83), confirming that YOLO-derived measurements captured essential dimensional variation. Inter-model comparisons revealed strong consistency (r ≥ 0.94, R² ≥ 0.96, MAE ≤ 0.40), indicating that YOLOv11 maintained the predictive reliability of YOLOv8 while operating with superior computational efficiency. This study establishes YOLOv11 as a robust foundation for AI-assisted digital breeding and automated quality monitoring systems in fungal research and precision agriculture.