As one of the world’s economically valuable fruit crops, citrus has its quality and productivity closely tied to the degree of fruit ripeness. However, accurately and efficiently detecting citrus ripeness in complex orchard environments for selective robotic harvesting remains a challenge. To address this, we constructed a citrus ripeness detection dataset under complex orchard conditions, proposed a lightweight algorithm based on visual saliency priors and the RT-DETR model, and named it LightSal-RTDETR. To reduce computational overhead, we designed the E-CSPPC module, which efficiently combines cross-stage partial networks with gated and partial convolutions, combined with cascaded group attention (CGA) and inverted residual mobile block (iRMB), which minimizes model complexity and computational demand and simultaneously strengthens the model’s capacity for feature representation. Additionally, the Inner-SIoU loss function was employed for bounding box regression, while a weight initialization method based on visual saliency maps was proposed. Experiments on our dataset show that LightSal-RTDETR achieves a mAP@50 of 81%, improving by 1.9% over the original model while reducing parameters by 28.1% and computational cost by 26.5%. Therefore, LightSal-RTDETR effectively solves the citrus ripeness detection problem in orchard scenes with high complexity, offering an efficient solution for smart agriculture applications.

Mandarin fruit detection provides crucial technical support for yield prediction and the precise identification and harvesting of mandarin fruits. However, challenges such as occlusion from leaves or branches, the presence of small or partially visible fruits, and limitations in model efficiency pose significant obstacles in a complex orchard environment. To tackle these issues, we propose ELD-YOLO, a lightweight detection framework designed to enhance edge detail preservation and improve the detection of small and occluded fruits. Our method incorporates edge-aware processing to strengthen feature representation, introduces a streamlined detection head that balances accuracy with computational cost, and employs an adaptive upsampling strategy to minimize information loss during feature scaling. Experiments on a mandarin fruit dataset show that ELD-YOLO achieves a precision of 89.7%, a recall of 83.7%, an mAP@50 of 92.1%, and an mAP@50:95 of 68.6% while reducing the parameter count by 15.4% compared with the baseline. These results demonstrate that ELD-YOLO provides an effective and efficient solution for fruit detection in complex orchard scenarios.

Precise fruit recognition is crucial for the automated picking of peaches. However, practical implementation encounters challenges, including high costs and low efficiency, which hinder the commercialization of picking robots. To tackle these challenges, this study establishes a synthetic peach dataset and introduces PeachYOLO, an efficient and lightweight model for peach object detection in complex orchard environments. Specifically, based on the lightweight object detection model You Only Look Once version 8 (YOLOv8), this study first replaces traditional convolutions in the detection head structure with Partial Convolution (PConv). This improvement reduces computational and memory requirements while effectively extracting spatial features. Secondly, within the feature output of the neck network, Deformable Convolutional Networks version 2 (DCNv2) is employed in place of traditional convolutions to improve the recognition of irregular targets. Finally, Coordinate Attention (CA) is integrated into the head network to focus precisely on essential image information. Experimental results demonstrate that PeachYOLO achieves a mAP of 93.8%, surpassing the original model by 1.0%. Furthermore, PeachYOLO’s computation is only 5.1 FLOPs (G), the number of parameters is 2.6M, and has an inference time of 1.9ms, which is a reduction of 37.0%, 13.6%, and 5.6%, respectively, compared with the original YOLOv8n algorithm. These results underscore the substantial improvements in detection speed, accuracy, and model size offered by PeachYOLO. Moreover, its suitability for peach detection in intricate orchard settings lays the groundwork for the realization of unmanned intelligent peach picking.

Monitoring ships on water surfaces encounters obstacles such as weather conditions, sunlight, and water ripples, posing significant challenges in accurately detecting target ships in real time. Synthetic Aperture Radar (SAR) offers a viable solution for real-time ship detection, unaffected by cloud coverage, precipitation, or light levels. However, SAR images are often affected by speckle noise, salt-and-pepper noise, and water surface ripple interference. This study introduces LCAS-DetNet, a Multi-Location Cross-Attention Ship Detection Network tailored for the ships in SAR images. Modeled on the YOLO architecture, LCAS-DetNet comprises a feature extractor, an intermediate layer (“Neck”), and a detection head. The feature extractor includes the computation of Multi-Location Cross-Attention (MLCA) for precise extraction of ship features at multiple scales. Incorporating both local and global branches, MLCA bolsters the network’s ability to discern spatial arrangements and identify targets via a cross-attention mechanism. Each branch utilizes Multi-Location Attention (MLA) and calculates pixel-level correlations in both channel and spatial dimensions, further combating the impact of salt-and-pepper noise on the distribution of objective ship pixels. The feature extractor integrates downsampling and MLCA stacking, enhanced with residual connections and Patch Embedding, to improve the network’s multi-scale spatial recognition capabilities. As the network deepens, we consider this structure to be cascaded and multi-scale, providing the network with a richer receptive field. Additionally, we introduce a loss function based on Wise-IoUv3 to address the influence of label quality on the gradient updates. The effectiveness of our network was validated on the HRSID and SSDD datasets, where it achieved state-of-the-art performance: a 96.59% precision on HRSID and 97.52% on SSDD.

Accurate high-throughput fruit detection is the core prerequisite for precision citrus management. Existing models face a critical trade-off between accuracy for small fruits and computational efficiency, restricting large-scale industry transformation. To resolve this, we propose Citrus-SGYOLOv2, an optimized deep learning architecture specifically engineered for high-throughput phenotypic monitoring. The primary contribution of this work lies in three synergistic innovations: a novel Symmetric Ghost Backbone that prunes architectural redundancy while maintaining hierarchical feature depth; a Citrus Color Prior Calibration Attention Mechanism (Citrus_SE) that embeds physiological chromaticity priors to suppress complex spectral noise from foliage; and a P2-layer-based full-scale fusion strategy designed to recover fine-grained spatial details lost during downsampling. Experiments on our self-built dataset show that Citrus-SGYOLOv2 achieves 95.54% mAP@50 and 77.13% mAP@50–95, outperforming YOLOv11s by 5.03 and 9.90 percentage points respectively. Notably, the model achieves a 48.8% reduction in parameters (4.84 M) while sustaining a high-throughput inference speed of 139.00 FPS. This research provides a robust and efficient foundational framework for intelligent yield estimation and precision orchard management.