A lightweight and enhanced model for detecting the Neotropical brown stink bug, Euschistus heros (Hemiptera: Pentatomidae) based on YOLOv8 for soybean fields

Ecological Informatics, Год журнала: 2024, Номер 80, С. 102543 - 102543

Опубликована: Фев. 27, 2024

Insect pest detection and monitoring are vital in an agricultural crop to help prevent losses be more precise sustainable regarding the consequent actions taken. Deep learning (DL) approaches have attracted attention, showing triumphant performance many image-based applications. In adult stage, this research considers detecting a insect soybean crops, Neotropical brown stink bug (Euschistus heros), from field images acquired by drones cellphones. We develop test improved YOLO-model convolutional neural network (CNN) with fewer parameters than other state-of-the-art models demonstrate its superior generalization average precision on public image datasets new data provided here. Considering proposal's time of response, possibility deploying technology for automatic management near future is promising. provide open code all experiments performed.

Язык: Английский

---

Advanced deep learning model for crop-specific and cross-crop pest identification

Expert Systems with Applications, Год журнала: 2025, Номер 274, С. 126896 - 126896

Опубликована: Фев. 24, 2025

Язык: Английский

---

Key Intelligent Pesticide Prescription Spraying Technologies for the Control of Pests, Diseases, and Weeds: A Review

Agriculture, Год журнала: 2025, Номер 15(1), С. 81 - 81

Опубликована: Янв. 1, 2025

In modern agriculture, plant protection is the key to ensuring crop health and improving yields. Intelligent pesticide prescription spraying (IPPS) technologies monitor, diagnose, make scientific decisions about pests, diseases, weeds; formulate personalized precision control plans; prevent pests through use of intelligent equipment. This study discusses IPSS from four perspectives: target information acquisition, processing, spraying, implementation control. acquisition section, identification based on images, remote sensing, acoustic waves, electronic nose are introduced. processing methods such as pre-processing, feature extraction, pest disease identification, bioinformatics analysis, time series data addressed. impact selection, dose calculation, time, method resulting effect formulation in a certain area explored. implement vehicle automatic technology, droplet characteristic technology their applications studied. addition, this future development prospectives IPPS technologies, including multifunctional systems, decision-support systems generative AI, sprayers. The advancement these will enhance agricultural productivity more efficient, environmentally sustainable manner.

Язык: Английский

---

A transformer-based model with feature compensation and local information enhancement for end-to-end pest detection

Computers and Electronics in Agriculture, Год журнала: 2025, Номер 231, С. 109920 - 109920

Опубликована: Янв. 22, 2025

Язык: Английский

---

DAMI-YOLOv8l: A multi-scale detection framework for light-trapping insect pest monitoring

Ecological Informatics, Год журнала: 2025, Номер unknown, С. 103067 - 103067

Опубликована: Фев. 1, 2025

Язык: Английский

---

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

Plants, Год журнала: 2024, Номер 13(5), С. 653 - 653

Опубликована: Фев. 27, 2024

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

Язык: Английский

---

An enhancement algorithm for head characteristics of caged chickens detection based on cyclic consistent migration neural network

Poultry Science, Год журнала: 2024, Номер 103(6), С. 103663 - 103663

Опубликована: Март 15, 2024

The enclosed multistory poultry housing is a type of enclosure widely used in industrial caged chicken breeding. Accurate identification and detection the comb eyes chickens farms using this can enhance managers' understanding health chickens. However, accuracy image will be affected by enclosure's entrance, which reduce precision. Therefore, paper proposes cage-gate removal algorithm based on big data deep learning Cyclic Consistent Migration Neural Network (CCMNN). method achieves automatic elimination restoration some key information through CCMNN network. Structural Similarity Index Measure (SSIM) between recovered original images test set 91.14%. Peak signal-to-noise ratio (PSNR) 25.34dB. To verify practicability proposed method, performance target analyzed both before after applying network detecting combs Different YOLOv8 algorithms, including YOLOv8s, YOLOv8n, YOLOv8m, YOLOv8x, were to paper. experimental results demonstrate that compared without processing, precision improved 11, 11.3, 12.8, 10.2%. Similarly, eye for 2.4, 10.2, 6.8, 9%. more complete outline obtained enhanced. These advancements offer valuable insights future researchers aiming deploy enhanced equipment, thereby contributing accurate assessment production farm conditions.

Язык: Английский

---

Quality Detection and Grading of Rose Tea Based on a Lightweight Model

Foods, Год журнала: 2024, Номер 13(8), С. 1179 - 1179

Опубликована: Апрель 12, 2024

Rose tea is a type of flower in China’s reprocessed category, which divided into seven grades, including super flower, primary bud, heart, yellow scattered and waste flower. Grading rose distinct quality levels practice that essential to boosting their competitive advantage. Manual grading inefficient. We provide lightweight model advance automation. Firstly, four kinds attention mechanisms were introduced the backbone compared. According experimental results, Convolutional Block Attention Module (CBAM) was chosen end due its ultimate capacity enhance overall detection performance model. Second, module C2fGhost utilized change original C2f neck lighten network while maintaining performance. Finally, we used SIoU loss place CIoU improve boundary regression The results showed mAP, precision (P), recall (R), FPS, GFLOPs, Params values proposed 86.16%, 89.77%, 83.01%, 166.58, 7.978, 2.746 M, respectively. Compared with model, P, R increased by 0.67%, 0.73%, 0.64%, GFLOPs decreased 0.88 0.411 respectively, speed comparable. this study also performed better than other advanced models. It provides theoretical research technical support for intelligent roses.

Язык: Английский

---

YOLO_MRC: A fast and lightweight model for real-time detection and individual counting of Tephritidae pests

Ecological Informatics, Год журнала: 2023, Номер 79, С. 102445 - 102445

Опубликована: Дек. 22, 2023

Tephritidae pests severely affect the quality and safety of various melons, fruits vegetable crops. However, many agricultural managers lack an adequate understanding level pest occurrence, resulting in misuse pesticides, which ultimately leads to environmental pollution economic loss. Therefore, real-time detection counting are important for timely spotting control. This work helps quickly determine distribution abundance current environment, thus providing data on conditions management optimize pesticide use. Nevertheless, fast speed, high accuracy, lightweight performance difficult balance. To address this problem, based YOLOv8n model, paper takes Bactrocera cucurbitae as target proposes a individual model pests, named YOLO_MRC. introduces three key improvements: (1) Constructing new module called Multicat into neck network increases focus by incorporating attention mechanism; (2) Reducing original heads two then adjusting their sizes decrease number parameters model; (3) Devising novel module, C2flite, enhance deep feature extraction capability backbone network. According above points, conducts ablation experiments compare performances different models. Experiments showed that significantly offsets large increase GFLOPs processing time caused reducing head can further reduce improve accuracy when combined with C2flite module. On our dataset, [email protected] YOLO_MRC reached 99.3%. Simultaneously, decreases 63.68%, is reduced 19.75%, shortened 5%. ensure validity compared four excellent models using manual results benchmark. achieves average 94%, demonstrating superior terms size time. explore multiclass insect tasks, we choose public dataset Pest_24_640 comparison state-of-the-art 3.6 ms 70.4% only 2.4 MB size, demonstrates potential detection.

Язык: Английский

---

Camouflage detection: Optimization-based computer vision for Alligator sinensis with low detectability in complex wild environments

Ecological Informatics, Год журнала: 2024, Номер 83, С. 102802 - 102802

Опубликована: Авг. 28, 2024

Alligator sinensis is an extremely rare species that possesses excellent camouflage, allowing it to fit perfectly into its natural environment. The use of camouflage makes detection difficult for both humans and automated systems, highlighting the importance modern technologies animal monitoring. To address this issue, we present YOLO v8-SIM, innovative technique specifically developed significantly enhance identification precision. v8-SIM utilizes a sophisticated dual-layer attention mechanism, optimized loss function called inner intersection-over-union (IoU), slim-neck cross-layer hopping. results our study demonstrate model achieves accuracy rate 91 %, recall 89.9 mean average precision (mAP) 92.3 % IoU threshold 0.5. In addition, operates at frame 72.21 frames per second (FPS) excels accurately recognizing objects are partially visible or smaller in size. further improve initiatives, suggest creating open-source collection data showcases A. native environment while using techniques. These developments collectively ability detect disguised animals, thereby promoting monitoring protection biodiversity, supporting ecosystem sustainability.

Язык: Английский

---