BiFPN-enhanced SwinDAT-based cherry variety classification with YOLOv8

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 13, 2025

Accurate classification of cherry varieties is crucial for their economic value and market differentiation, yet genetic diversity visual similarity make manual identification challenging, hindering efficient agricultural trade practices. This study addresses this issue by proposing a novel deep learning-based hybrid model that integrates BiFPN with the YOLOv8n-cls framework, enhanced Swin Transformer Deformable Attention (DAT) techniques. The was trained evaluated on newly constructed dataset comprising from Turkey's Western Mediterranean region. Experimental results demonstrated effectiveness proposed approach, achieving precision 91.91%, recall 92.0%, F1-score 91.93%, an overall accuracy 91.714%. findings highlight model's potential to optimize harvest timing, ensure quality control, support export classification, thereby contributing improved practices outcomes.

Language: Английский

---

A lightweight deep learning framework for wild berry detection in complex natural environments

Engineering Applications of Artificial Intelligence, Journal Year: 2025, Volume and Issue: 154, P. 110918 - 110918

Published: April 29, 2025

Language: Английский

---

Experimental analysis and point cloud-based prediction approach for axial compression behavior of randomly corroded steel tubes

Thin-Walled Structures, Journal Year: 2025, Volume and Issue: unknown, P. 113498 - 113498

Published: May 1, 2025

Language: Английский

---

A Swin transformer and MLP based method for identifying cherry ripeness and decay

Frontiers in Physics, Journal Year: 2023, Volume and Issue: 11

Published: Oct. 20, 2023

Cherries are a nutritionally beneficial and economically significant crop, with fruit ripeness decay (rot or rupture) being critical indicators in the cherry sorting process. Therefore, accurately identifying maturity of cherries is crucial processing. With advancements artificial intelligence technology, many studies have utilized photographs for non-destructive detection appearance quality. This paper proposes quality identification method based on Swin Transformer, which utilizes Transformer to extract image feature information then imports into classifiers such as multi-layer perceptron(MLP) support vector machine(SVM) classification. Through comparison multiple classifiers, optimal classifier, namely, MLP, combination obtained. Furthermore, performance comparisons conducted original Swin-T method, traditional CNN models, models combined MLP. The results demonstrate following: 1) proposed MLP achieves an accuracy rate 98.5%, 2.1% higher than model 1.0% best-performing 2) training time required only 78.43 s, significantly faster other models. experimental indicate that innovative approach combining shows excellent decay. successful application this provides new solution determining plays role promoting development machines.

Language: Английский

---

Research on the Strawberry Recognition Algorithm Based on Deep Learning

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(20), P. 11298 - 11298

Published: Oct. 14, 2023

In view of the time-consuming and laborious manual picking sorting strawberries, direct impact image recognition accuracy on automatic rapid development deep learning(DL), a Faster Regions with Convolutional Neural Network features (R-CNN) strawberry method that combines Mixup data augmentation, ResNet(Residual Network)50 backbone feature extraction network Soft-NMS (Non-Maximum Suppression) algorithm, named MRS R-CNN, is proposed. this paper, transfer learning VGG (Visual Geometry Group) 16 ResNet50 are compared, superior selected as R-CNN. The augmentation fusion used to improve generalization ability model. redundant bboxes (bounding boxes) removed through obtain best region proposal. freezing phase added training process, effectively reducing occupation video memory shortening time. After experimental verification, optimized model improved AP (Average Precision) values mature immature strawberries by 0.26% 5.34%, respectively, P(Precision) 0.81% 6.34%, compared original (R R-CNN). Therefore, R-CNN proposed in paper has great potential field maturity classification improves rate small fruit overlapping occluded fruit, thus providing an excellent solution for mechanized sorting.

Language: Английский

---

Identifying cherry maturity and disease using different fusions of deep features and classifiers

Journal of Food Measurement & Characterization, Journal Year: 2023, Volume and Issue: 17(6), P. 5794 - 5805

Published: Aug. 3, 2023

Language: Английский

---

Impact of Digital Innovation on Corporate Productivity: A Predictive Model Based on Data Mining

Published: May 30, 2024

Language: Английский

---

ST-MobileNetV3: A Lightweight Network Model for Strawberry Disease Identification

2022 International Conference on Networking and Network Applications (NaNA), Journal Year: 2023, Volume and Issue: unknown, P. 209 - 214

Published: Aug. 1, 2023

The identification of strawberry diseases holds great significance in the cultivation process, and timely detection plays a vital role promoting production advancing industry. This paper introduces novel disease recognition network model named ST-MobileNetV3. Building upon MobileNetV3, incorporates multilayer perception module expands convolutional processing, while replacing original attention mechanism SE with an ECA module. research achieves harmonious balance between accuracy complexity, thereby supporting development lightweight techniques. introduction this innovative is expected to offer efficient accurate tools growers, facilitating progress

Language: Английский

---