达 幸 杨

Mr. 达 幸 杨


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 December 2025 in Electronics
Verified
Journal Article
Open access
PDF file
Liang YuLiang Yu
Liang Yu
XFXingcan Fang
Xingcan Fang
YZYuze Zeng
Yuze Zeng
WGWeili Guo
Weili Guo
达杨达 幸 杨
达 幸 杨
HSHengyi Sun
Hengyi Sun
The automation of quality control in agricultural food processing, particularly the detection of incomplete peeling in taro, constitutes a critical frontier for ensuring food safety and optimizing production efficiency in the Industry 4.0 era. However, this domain is fraught with significant technical challenges, primarily stemming from the inherent visual characteristics of residual peel: extremely minute scales relative to the vegetable body, highly irregular morphological variations, and the dense occlusion of objects on industrial conveyor belts. To address these persistent impediments, this study introduces a comprehensive solution comprising a specialized dataset and a novel detection architecture. We established the Taro Peel Industrial Dataset (TPID), a rigorously annotated collection of 18,341 high-density instances reflecting real-world production conditions. Building upon this foundation, we propose MDB-YOLO, a lightweight, multi-dimensional bionic detection model evolved from the YOLOv8s architecture. The MDB-YOLO framework integrates a synergistic set of innovations designed to resolve specific detection bottlenecks. To mitigate the conflict between background texture interference and tiny target detection, we integrated the C2f_EMA module with a Wise-IoU (WIoU) loss function, a combination that significantly enhances feature response to low-contrast residues while reducing the penalty on low-quality anchor boxes through a dynamic non-monotonic focusing mechanism. To effectively manage irregular peel shapes, a dynamic feature processing chain was constructed utilizing DySample for morphology-aware upsampling, BiFPN_Concat2 for weighted multi-scale fusion, and ODConv2d for geometric preservation. Furthermore, to address the issue of missed detections caused by dense occlusion in industrial stacking scenarios, Soft-NMS was implemented to replace traditional greedy suppression mechanisms. Experimental validation demonstrates the superiority of the proposed framework. MDB-YOLO achieves a mean Average Precision (mAP50-95) of 69.7% and a Recall of 88.0%, significantly outperforming the baseline YOLOv8s and advanced transformer-based models like RT-DETR-L. Crucially, the model maintains high operational efficiency, achieving an inference speed of 1.1 ms on an NVIDIA A100 and reaching 27 FPS on an NVIDIA Jetson Xavier NX using INT8 quantization. These findings confirm that MDB-YOLO provides a robust, high-precision, and cost-effective solution for real-time quality control in agricultural food processing, marking a significant advancement in the application of computer vision to complex biological targets.

15 August 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
Liang YuLiang Yu
Liang Yu
XFXingcan Fang
Xingcan Fang
YZYuze Zeng
Yuze Zeng
WGWeili Guo
Weili Guo
达杨达 幸 杨
达 幸 杨
HSHengyi Sun
Hengyi Sun
The automation of quality control in the food processing industry is critical for enhancing production efficiency, ensuring product safety, and reducing operational costs. A significant challenge in this domain is the automated inspection of tuber peeling, where incomplete processing can leave undesirable residual fragments. This paper addresses the specific problem of detecting small, low-contrast, and densely clustered residual taro peel fragments in a real-time industrial setting. We propose Multi-Dimensional Bionic YOLO (MDB-YOLO), a novel, lightweight object detection model based on the YOLOv8 architecture, specifically optimized for this challenging task. MDB-YOLO integrates several key innovations to achieve a superior balance of accuracy and efficiency: a C2f_EMA module enhances multi-scale feature representation through an efficient attention mechanism; Dynamic Upsampling (DySample) improves the reconstruction of high-resolution details; Omni-Dimensional Dynamic Convolution (ODConv2d) provides adaptive kernel learning; a BiFPN_Concat2 module is used for feature fusion; the Wise-IoU (WIoU) loss function focuses training on difficult examples; and Soft-NMS improves recall in dense scenes. Evaluated on a custom-built Taro Peel Industrial Dataset (TPID), MDB-YOLO achieves a mean Average Precision (mAP@0.5) of 92.1% and a mAP@0.5:0.95 of 69.7%, outperforming the YOLOv8s baseline and other state-of-the-art models. With only 13.44 million parameters and a computational load of 28.4 GFLOPS, MDB-YOLO demonstrates its suitability for lightweight deployment on resource-constrained edge devices, offering a practical and effective solution for automated quality control by delivering top-tier accuracy at a fraction of the computational cost and inference time of its competitors.