Bark beetle (Dendroctonus spp) is one of the typical wood-boring pests. It has posed a serious threat to the forest resource, due to the small size, cryptic behavior, and long damage cycles. According to the monitoring data from the National Forestry and Grassland Administration, the infestations of these pests occurred in 2024 in southwestern, northern, and northwestern China. There is the infectious area of 184 000 hectares, with the moderate to severe damage accounting for 15.08% of the total. Yet, their control and prevention strategies can differ substantially, due to the different species of the bark beetle with the sympatric distribution. For instance, the Dendroctonus micansrequires removal of the infested trees is utilized with γ-hexachlorocyclohexane (Lindane) treatment; controlling Dendroctonus valensinvolves is often combined with adult eradication and aluminum phosphide; whereas Heterobostrychus hamatipennisrelies can depend on methyl bromide or aluminum phosphide fumigation. Moreover, the interspecies similarity in macroscopic features (such as the body length and coloration) can cause the identification to be highly dependent on the local microscopic features, including the shape of the disc and elytral punctures. The typical fine-grained recognition is often required to accurately distinguish different species within the highly similar base category (Scolytidae), according to the subtle discriminative features. Furthermore, manual identification of the morphologies is highly subjective; In this study, a rapid and accurate fine-grained recognition was developed to prevent and control the bark beetles. A FGRS-Net (Fine-Grained Recognition for Scolytidae Network) architecture was also constructed to identify the bark beetles. Multi-level technologies were proposed to systematically solve the key issues in bark beetle recognition, including the scale variation, feature confusion, and computational efficiency. Firstly, a detection head module with multi-modal embedding was proposed to mitigate the inter-class recognition bias caused by insufficient training samples. Morphological feature vectors, local texture descriptors, and spatial contextual information were integrated to significantly reduce the false detection rates induced by uneven sample distribution. A joint embedding space was then constructed to effectively enhance the discrimination for the morphologically similar species. Secondly, an Attention Convolution Mixer (ACmix) module was introduced for the large size range and variable habitat postures of the bark beetles. The multi-scale receptive fields were adaptively adjusted using the parallel convolutional paths and self-attention mechanisms. This module was realized to capture the local details of the millimeter-scale pests (such as elytral punctures and antenna morphology). While the overall distribution patterns were effectively identified in the aggregated populations. Thereby, the robustness of the feature discrimination was improved in complex backgrounds. An Omni-Dimensional Dynamic Convolution (ODConv) module was integrated to further optimize the feature representation efficiency. A four-dimensional attention mechanism was constructed (across spatial, channel, kernel, and network depth dimensions). The dynamic generation and adaptive calibration of the convolutional parameters significantly reduced the number of parameters. While the key discriminative features were enhanced, such as the wing venation structure and body segment proportions. In model lightweighting, a combined optimization was adopted on structured pruning and knowledge distillation. Channel importance was constrained via L1 regularization to prune the redundant feature connections. While a multi-teacher distillation framework was designed to transfer the hierarchical feature representations from large networks to a lightweight student model. As a result, the model size was compressed by 40.7%, and the inference latency was reduced by 35%, indicating the high accuracy. A multi-interference condition testing system was constructed to validate the applicability in practical scenarios. Complex field environments were simulated, including lens fog, low illumination, blur, and foliage occlusion. Deployment verification was conducted on the edge devices with different computational architectures. Experimental results show that the FGRS-Net achieved a mean average precision (mAP) of 89.3% and a recall rate of 98% on the self-built fine-grained bark beetle dataset, with a 16% reduction in the floating point operations (FLOPs) and a detection speed of 289 FPS. In edge device deployment, the Raspberry Pi platform achieved real-time inference at 11 FPS, while the RK3576 platform reached a processing speed of 27 FPS. The technical solution can provide reliable technical support for accurate monitoring of bark beetles in field environments. The finding can offer important references for the pest recognition models in the field of smart forestry.

Rice is one of the most important staple crops worldwide. Its yield and quality can directly influence the global food security and the agricultural economy. Nevertheless, the rice pests have ranked the most among common biological disasters that threaten stable, high yield rice production. The International Rice Research Institute has reported that the rice pests and diseases have cut the yields by up to 37%, with the losses ranging from 24% to 41%. Real-time monitoring and counting of pests are often required for the prevention and control in green agriculture. It is very necessary for the accurate identification of pests at different scales and reliable tracking of their population dynamics Yet the pest community in paddy fields is extraordinarily diverse. Pest sizes are often ranged from the millimeter scale aphids and thrips to stem borers and leaf folder larvae exceeding ten millimeter. All pest can concurrently occur in the same plot, and then simultaneously inhabit leaves, leaf sheaths, stems, or panicles. Conventional manual scouting or simple image processing cannot fully meet the accurate detection and counting, due to the complex spatial distribution, extreme morphological variation, and heavy background clutter. In this study, the YOLO-MSLP (multi-scale lightweight pest), an intelligent lightweight model was proposed for the rice-pest detection and counting, in order to overcome these challenges. The latest YOLOv11n backbone, YOLO-MSLP was introduced three innovations that tailored to the complex scenes in the paddy field. Firstly, an adaptive pooling bidirectional feature pyramid network (AP-BiFPN) was embedded in the neck. The adaptive pooling was dynamically adjusted the receptive field and bidirectional cross scale fusion. Multiscale features were extracted and then aggregated in a stable manner, whether the targets were solitary pests or dense clusters. The small object detection was greatly improved for the accuracy of the large object localization. Secondly, a multi-scale triplet attention module (MS-TAM) was inserted between the backbone and detection heads. The channel, spatial, and scale dimensions were operated in parallel. The discriminative pest features were adaptively highlighted to suppress the redundant background information closely resembled the pests, such as the shape, texture, and color. Experimental results showed that the module was maintained on the high confidence outputs even under back lighting, leaf occlusion, or overlapping rice plants. Finally, the backbone was reengineered with a reparametrized vision transformer (RepViT), in order to lower deployment barriers. Furthermore, knowledge distillation was compressed to transfer the rich representations from a larger teacher network into the lightweight student. The YOLO-MSLP was achieved a mean average precision (mAP) of 94.5% and a recall of 91.7%, after pruning, quantization, and operator fusion. Floating point operations were reduced by 24.4%, and model size was shrunk by 40.7%. Inference latency for a single image on an edge GPU fell below 35 ms. Extensive testing confirmed that the YOLO MSLP can run in real time on embedded devices, thus providing for a low-cost, highly reliable tool for early warning, precise spraying, and green control of rice pests. The model can be expected for the large-scale smart-agriculture deployments to advance the sustainable rice industry. The finding can also provide the data referenec for the scientific interventions, thereby reducing the pesticide use and residue risk.