UAV small target detection algorithm based on an improved YOLOv5s model

Journal of Visual Communication and Image Representation, Год журнала: 2023, Номер 97, С. 103936 - 103936

Опубликована: Сен. 21, 2023

Язык: Английский

---

A lightweight SOD-YOLOv5n model-based winter jujube detection and counting method deployed on Android

Computers and Electronics in Agriculture, Год журнала: 2024, Номер 218, С. 108701 - 108701

Опубликована: Фев. 13, 2024

Язык: Английский

---

EFLNet: Enhancing Feature Learning Network for Infrared Small Target Detection

IEEE Transactions on Geoscience and Remote Sensing, Год журнала: 2024, Номер 62, С. 1 - 11

Опубликована: Янв. 1, 2024

Язык: Английский

---

Infrared small target detection with super-resolution and YOLO

Optics & Laser Technology, Год журнала: 2024, Номер 177, С. 111221 - 111221

Опубликована: Май 24, 2024

Язык: Английский

---

Enhanced-YOLOv8: A new small target detection model

Digital Signal Processing, Год журнала: 2024, Номер 153, С. 104611 - 104611

Опубликована: Май 31, 2024

Язык: Английский

---

YOLO-MST: Multiscale deep learning method for infrared small target detection based on super-resolution and YOLO

Optics & Laser Technology, Год журнала: 2025, Номер 187, С. 112835 - 112835

Опубликована: Март 31, 2025

Язык: Английский

---

ISTD-DETR: A deep learning algorithm based on DETR and Super-resolution for infrared small target detection

Neurocomputing, Год журнала: 2025, Номер unknown, С. 129289 - 129289

Опубликована: Янв. 1, 2025

Язык: Английский

---

Fine-grained mineral recognition and segmentation in metal mineral optical images based on improved YOLOv8n model

Minerals Engineering, Год журнала: 2025, Номер 227, С. 109290 - 109290

Опубликована: Апрель 8, 2025

Язык: Английский

---

CA-YOLO: Model Optimization for Remote Sensing Image Object Detection

IEEE Access, Год журнала: 2023, Номер 11, С. 64769 - 64781

Опубликована: Янв. 1, 2023

The CA-YOLO (Coordinate Attention-YOLO) model has been optimized for object detection in complex remote sensing images, addressing key issues faced by multi-object algorithms. These include weak multi-scale feature learning capabilities and the challenging tradeoff between accuracy parameter complexity. Built on framework of YOLOv5, incorporates a lightweight coordinate attention module shallow layer to improve detailed extraction suppress redundant information interference, while spatial pyramid pooling-fast with tandem construction is implemented deeper layer. also employs stochastic pooling strategy fuse from low-level high-level layers, reducing number parameters improving inference speed. anchor box mechanism modified loss function have enhance learning. Results show that outperforms original YOLO terms accuracy, an average [email protected] improvement 4.8% [email protected]:0.95 3.8%. Additionally, demonstrates exceptional speed, averaging 125 fps, which reinforces its superiority generalization ability, overall efficiency. Notably, these improvements were achieved maintaining same complexity as other models, making choice various applications.

Язык: Английский

---

Detection of the Pine Wilt Disease Using a Joint Deep Object Detection Model Based on Drone Remote Sensing Data

Forests, Год журнала: 2024, Номер 15(5), С. 869 - 869

Опубликована: Май 16, 2024

Disease and detection is crucial for the protection of forest growth, reproduction, biodiversity. Traditional methods face challenges such as limited coverage, excessive time resource consumption, poor accuracy, diminishing effectiveness disease prevention control. By addressing these challenges, this study leverages drone remote sensing data combined with deep object models, specifically employing YOLO-v3 algorithm based on loss function optimization, efficient accurate tree diseases pests. Utilizing drone-mounted cameras, captures insect pest image information in pine areas, followed by segmentation, merging, feature extraction processing. The computing system airborne embedded devices designed to ensure efficiency accuracy. improved CIoU was used detect pests diseases. Compared traditional IoU function, takes into account overlap area, distance between center predicted frame actual frame, consistency aspect ratio. experimental results demonstrate proposed model’s capability process images at a slightly faster speed, an average processing less than 0.5 s per image, while achieving accuracy surpassing 95%. identifying high comprehensiveness offers significant potential developing inspection plans. However, limitations exist performance complex environments, necessitating further research improve model universality adaptability across diverse regions. Future directions include exploring advanced models minimize demands enhance practical application support

Язык: Английский

---