An Efficient Pest Detection Framework with a Medium-Scale Benchmark to Increase the Agricultural Productivity

Image detection model construction of Apolygus lucorum and Empoasca spp. based on improved YOLOv5

BACKGROUND

The polyphagous mirid bug Apolygus lucorum (Meyer-Dür) and the green leafhopper Empoasca spp. Walsh are small pests that are widely distributed and important pests of many economically important crops, especially kiwis. Conventional monitoring methods are expensive, laborious and error-prone. Currently, deep learning methods are ineffective at recognizing them. This study proposes a new deep-learning-based YOLOv5s_HSSE model to automatically detect and count them on sticky card traps.

RESULTS

Based on a database of 1502 images, all images were collected from kiwi orchards at multiple locations and times. We trained the YOLOv5s model to detect and count them and then changed the activation function to Hard swish in YOLOv5s, introduced the SIoU Loss function, and added the squeeze-and-excitation attention mechanism to form a new YOLOv5s_HSSE model. Mean average precision of this model in the test dataset was 95.9%, the recall rate was 93.9% and the frames per second was 155, which are higher than those of other single-stage deep-learning models, such as SSD, YOLOv3 and YOLOv4.

CONCLUSION

The proposed YOLOv5s_HSSE model can be used to identify and count A. lucorum and Empoasca spp., and it is a new, efficient and accurate monitoring method. Pest detection will benefit from the broader applications of deep learning. © 2024 Society of Chemical Industry.

AI-Enabled Crop Management Framework for Pest Detection Using Visual Sensor Data

Our research focuses on addressing the challenge of crop diseases and pest infestations in agriculture by utilizing UAV technology for improved crop monitoring through unmanned aerial vehicles (UAVs) and enhancing the detection and classification of agricultural pests. Traditional approaches often require arduous manual feature extraction or computationally demanding deep learning (DL) techniques. To address this, we introduce an optimized model tailored specifically for UAV-based applications. Our alterations to the YOLOv5s model, which include advanced attention modules, expanded cross-stage partial network (CSP) modules, and refined multiscale feature extraction mechanisms, enable precise pest detection and classification. Inspired by the efficiency and versatility of UAVs, our study strives to revolutionize pest management in sustainable agriculture while also detecting and preventing crop diseases. We conducted rigorous testing on a medium-scale dataset, identifying five agricultural pests, namely ants, grasshoppers, palm weevils, shield bugs, and wasps. Our comprehensive experimental analysis showcases superior performance compared to various YOLOv5 model versions. The proposed model obtained higher performance, with an average precision of 96.0%, an average recall of 93.0%, and a mean average precision (mAP) of 95.0%. Furthermore, the inherent capabilities of UAVs, combined with the YOLOv5s model tested here, could offer a reliable solution for real-time pest detection, demonstrating significant potential to optimize and improve agricultural production within a drone-centric ecosystem.

Yolo-Pest: An Insect Pest Object Detection Algorithm via CAC3 Module

Insect pests have always been one of the main hazards affecting crop yield and quality in traditional agriculture. An accurate and timely pest detection algorithm is essential for effective pest control; however, the existing approach suffers from a sharp performance drop when it comes to the pest detection task due to the lack of learning samples and models for small pest detection. In this paper, we explore and study the improvement methods of convolutional neural network (CNN) models on the Teddy Cup pest dataset and further propose a lightweight and effective agricultural pest detection method for small target pests, named Yolo-Pest, for the pest detection task in agriculture. Specifically, we tackle the problem of feature extraction in small sample learning with the proposed CAC3 module, which is built in a stacking residual structure based on the standard BottleNeck module. By applying a ConvNext module based on the vision transformer (ViT), the proposed method achieves effective feature extraction while keeping a lightweight network. Comparative experiments prove the effectiveness of our approach. Our proposal achieves 91.9% mAP0.5 on the Teddy Cup pest dataset, which outperforms the Yolov5s model by nearly 8% in mAP0.5. It also achieves great performance on public datasets, such as IP102, with a great reduction in the number of parameters.