Classification and detection of insects from field images using deep learning for smart pest management: A systematic review

Ecological Informatics, Journal Year: 2021, Volume and Issue: 66, P. 101460 - 101460

Published: Oct. 14, 2021

Language: Английский

---

The impact of climate change on insect pest biology and ecology: Implications for pest management strategies, crop production, and food security

Journal of Agriculture and Food Research, Journal Year: 2023, Volume and Issue: 14, P. 100733 - 100733

Published: Aug. 9, 2023

The explosive expansion of the global population and technological progress has greatly influenced agriculture food production. However, this is threatened by climate change, which unleashes a slew issues like carbon dioxide (CO2) increases, frequent droughts, temperature shifts that present substantial obstacle to crop yields security. ramifications these climatic factors on insect pest biology ecology are profound, given pests depend heavily factors. Since productivity tightly connected both variables, changes in can significantly impact yields. Therefore, it imperative comprehend change manage them effectively ensure sufficient This review examines effect explores potential use modern monitoring technologies prediction tools devise effective management strategies improve production

Language: Английский

---

Recent Trends in Internet-of-Things-Enabled Sensor Technologies for Smart Agriculture

IEEE Internet of Things Journal, Journal Year: 2022, Volume and Issue: 9(23), P. 23583 - 23598

Published: Sept. 27, 2022

Smart agriculture integrates key information communication technologies with sensing to provide effective and cost-efficient agricultural services. leverages a wide range of advanced technologies, such as wireless sensor networks, Internet Things, robotics, bots, drones, artificial intelligence, cloud computing. The adoption these in smart enables all stakeholders the sector develop better managerial decisions get more yield. We differentiate between traditional based on deployment architectures along focus various processing stages agriculture. present comprehensive review types sensors that are playing vital role enabling also integration emerging computing infrastructures make smarter. Finally, we discuss open research challenges must be addressed improve future.

Language: Английский

---

Accurate detection and identification of insects from camera trap images with deep learning

PLOS Sustainability and Transformation, Journal Year: 2023, Volume and Issue: 2(3), P. e0000051 - e0000051

Published: March 15, 2023

Reported insect declines have dramatically increased the global demand for standardized monitoring data. Image-based can generate such data cost-efficiently and non-invasively. However, extracting ecological from images is more challenging insects than vertebrates because of their small size great diversity. Deep learning facilitates fast accurate detection identification, but lack training coveted deep models a major obstacle application. We present large annotated image dataset functionally important taxa. The primary consists 29,960 representing nine taxa including bees, hoverflies, butterflies beetles across two million recorded with ten time-lapse cameras mounted over flowers during summer 2019. was extracted using an iterative approach: First, preliminary model identified candidate insects. Second, were manually screened by users online citizen science platform. Finally, all annotations quality checked experts. used to train compare performance selected You Only Look Once (YOLO) algorithms. show that these detect classify in complex scenes unprecedented accuracy. best performing YOLOv5 consistently identifies dominant species play roles pollination pest control Europe. reached average precision 92.7% recall 93.8% classification species. Importantly, when presented uncommon or unclear not seen training, our detects 80% individuals usually interprets them as closely related This useful property (1) rare which are absent, (2) new correctly identify those future. Our camera system, framework promising results non-destructive Furthermore, resulting quantify phenology, abundance, foraging behaviour flower-visiting Above all, this represents critical first benchmark future development evaluation identification.

Language: Английский

---

A Systematic Review on Automatic Insect Detection Using Deep Learning

Agriculture, Journal Year: 2023, Volume and Issue: 13(3), P. 713 - 713

Published: March 19, 2023

Globally, insect pests are the primary reason for reduced crop yield and quality. Although pesticides commonly used to control eliminate these pests, they can have adverse effects on environment, human health, natural resources. As an alternative, integrated pest management has been devised enhance control, decrease excessive use of pesticides, output quality crops. With improvements in artificial intelligence technologies, several applications emerged agricultural context, including automatic detection, monitoring, identification insects. The purpose this article is outline leading techniques automated detection insects, highlighting most successful approaches methodologies while also drawing attention remaining challenges gaps area. aim furnish reader with overview major developments field. This study analysed 92 studies published between 2016 2022 insects traps using deep learning techniques. search was conducted six electronic databases, 36 articles met inclusion criteria. criteria were that applied classification, counting, written English. selection process involved analysing title, keywords, abstract each study, resulting exclusion 33 articles. included 12 classification task 24 task. Two main approaches—standard adaptable—for identified, various architectures detectors. accuracy found be influenced by dataset size, significantly affected number classes size. highlights two recommendations, namely, characteristics (such as unbalanced incomplete annotation) limitations algorithms small objects lack information about insects). To overcome challenges, further research recommended improve practices. should focus addressing identified ensure more effective management.

Language: Английский

---

PICT: A low‐cost, modular, open‐source camera trap system to study plant–insect interactions

Methods in Ecology and Evolution, Journal Year: 2021, Volume and Issue: 12(8), P. 1389 - 1396

Published: April 25, 2021

Abstract Commercial camera traps (CTs) commonly used in wildlife studies have several technical limitations that restrict their scope of application. They are not easily customizable, unit prices sharply increase with image quality and importantly, they designed to record the activity ectotherms such as insects. Those developed for study plant–insect interactions yet be widely adopted rely on expensive heavy equipment. We PICT (plant–insect trap), an inexpensive (<100 USD) do‐it‐yourself CT system based a Raspberry Pi Zero computer continuously film animal activity. The is particularly well suited pollination, insect behaviour predator–prey interactions. focus distance can manually adjusted under 5 cm. In low light conditions, near‐infrared automatically illuminates subject. Frame rate, resolution video compression levels set by user. remotely controlled using either smartphone, tablet or laptop via onboard Wi‐Fi. up 72‐hr day night videos at >720p 110‐Wh power bank (30,000 mAh). Its ultra‐portable (<1 kg) waterproof design modular architecture practical diverse field settings. provide illustrated guide detailing steps involved building operating post‐processing. successfully field‐tested Central African rainforest two contrasting research settings: pollinator survey canopy ebony Diospyros crassiflora observation rare pollination events epiphytic orchid Cyrtorchis letouzeyi . overcomes many associated systems monitor ectotherms. Increased portability lower costs allow large‐scale deployment acquisition novel insights into reproductive biology plants difficult observe animals. ​

Language: Английский

---

Explainable deep convolutional neural networks for insect pest recognition

Journal of Cleaner Production, Journal Year: 2022, Volume and Issue: 371, P. 133638 - 133638

Published: Aug. 20, 2022

Language: Английский

---

Forest Insect Biosecurity: Processes, Patterns, Predictions, Pitfalls

Annual Review of Entomology, Journal Year: 2022, Volume and Issue: 68(1), P. 211 - 229

Published: Oct. 6, 2022

The economic and environmental threats posed by non-native forest insects are ever increasing with the continuing globalization of trade travel; thus, need for mitigation through effective biosecurity is greater than ever. However, despite decades research implementation preborder, border, postborder preventative measures, insect invasions continue to occur, no evidence saturation, even predicted accelerate. In this article, we review measures used mitigate arrival, establishment, spread, impacts possible impediments successful these measures. Biosecurity successes likely under-recognized because they difficult detect quantify, whereas failures more evident in continued establishment additional species. There limitations existing systems at global country scales (for example, inspecting all imports impossible, phytosanitary perfect, knownunknowns cannot be regulated against, noncompliance an ongoing problem). should a shared responsibility across countries, governments, stakeholders, individuals.

Language: Английский

---

Environmental DNA metabarcoding from flowers reveals arthropod pollinators, plant pests, parasites, and potential predator–prey interactions while revealing more arthropod diversity than camera traps

Environmental DNA, Journal Year: 2023, Volume and Issue: 5(3), P. 551 - 569

Published: March 29, 2023

Abstract Arthropods can strongly impact ecosystems through pollination, herbivory, predation, and parasitism. As such, characterizing arthropod biodiversity is vital to understanding ecosystem health, functions, services. Emerging environmental DNA (eDNA) methods targeting trace eDNA left behind on flowers have the potential track interactions. The goal of this study was determine extent which metabarcoding identify plant‐arthropod arthropod‐arthropod interactions assess compared conventional sampling. We deployed camera traps document activity specific flowers, sampled from those same then performed a analysis that targets partial fragment cytochrome c oxidase subunit I gene (COI) all present. found our detected small pollinators, plant pests, parasites, shed light predator–prey while detecting 55 species just 21 trapping. trapping survey, however, larger, more conspicuous nectarivores successfully. also explored ecology residual eDNA, finding rainfall had significant negative effect ability detect eDNA. Preliminary evidence indicates flower may amount be detected. provide clues highlights insights gained future studies. show valuable tool for not only pollinator communities but revealing among plants, predators. Future research should focus how improve detection large pollinators/nectivores studying further explore method's utility.

Language: Английский

---

Hierarchical classification of insects with multitask learning and anomaly detection

Ecological Informatics, Journal Year: 2023, Volume and Issue: 77, P. 102278 - 102278

Published: Aug. 28, 2023

Cameras and computer vision are revolutionising the study of insects, creating new research opportunities within agriculture, epidemiology, evolution, ecology monitoring biodiversity. However, diversity insects close resemblances many species a major challenge for image-based species-level classification. Here, we present an algorithm to hierarchically classify from images, leveraging simple taxonomy (1) specimens across multiple taxonomic ranks simultaneously, (2) identify lowest rank at which reliable classification can be reached. Specifically, propose multitask learning, loss function incorporating class dependency each rank, anomaly detection based on outlier analysis quantify uncertainty. First, compile dataset 41,731 images combining time-lapse floral scenes with Global Biodiversity Information Facility (GBIF). Second, adapt state-of-the-art convolutional neural networks, ResNet EfficientNet, hierarchical belonging three orders, five families nine species. Third, assess model generalization 11 unseen by trained models. is used predict higher were not in training set. We found that into our increased accuracy ranks. As expected, correctly classified insect ranks, while was uncertain lower Anomaly effectively flag novel taxa visually distinct data. consistently mistaken similar Above all, have demonstrated practical approach uncertainty during automated situ live insects. Our method versatile, forming valuable step towards high-level

Language: Английский

---