Detection and classification of soybean pests using deep learning with UAV images

Tetila, Everton Castelão; Machado, Bruno Brandoli; Astolfi, Gilberto; Belete, Nícolas Alessandro de Souza; Amorim, Willian Paraguassu; Roel, Antônia Railda; Pistori, Hemerson

doi:10.1016/j.compag.2020.105836

Cited by 169 publications

(92 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…evaluated [31]. The results of this study showed that with transfer learning and fine-tuning techniques, very good accuracy can be achieved.…”

Section: Location and Description Of The Study Areamentioning

confidence: 72%

“…These researchers claimed their method had advantages such as better robustness, generalization and acceptable accuracy. In another study, different models of deep networks for classification of soybean pest images were evaluated [31]. The results of this study showed that with transfer learning and fine-tuning techniques, very good accuracy can be achieved.…”

Section: Deep Learningmentioning

confidence: 97%

See 1 more Smart Citation

A Novel Technique for Classifying Bird Damage to Rapeseed Plants Based on a Deep Learning Algorithm

et al. 2021

View full text Add to dashboard Cite

Estimation of crop damage plays a vital role in the management of fields in the agriculture sector. An accurate measure of it provides key guidance to support agricultural decision-making systems. The objective of the study was to propose a novel technique for classifying damaged crops based on a state-of-the-art deep learning algorithm. To this end, a dataset of rapeseed field images was gathered from the field after birds’ attacks. The dataset consisted of three classes including undamaged, partially damaged, and fully damaged crops. Vgg16 and Res-Net50 as pre-trained deep convolutional neural networks were used to classify these classes. The overall classification accuracy reached 93.7% and 98.2% for the Vgg16 and the ResNet50 algorithms, respectively. The results indicated that a deep neural network has a high ability in distinguishing and categorizing different image-based datasets of rapeseed. The findings also revealed a great potential of deep learning-based models to classify other damaged crops.

show abstract

“…evaluated [31]. The results of this study showed that with transfer learning and fine-tuning techniques, very good accuracy can be achieved.…”

Section: Location and Description Of The Study Areamentioning

confidence: 72%

Section: Deep Learningmentioning

confidence: 97%

A Novel Technique for Classifying Bird Damage to Rapeseed Plants Based on a Deep Learning Algorithm

et al. 2021

View full text Add to dashboard Cite

show abstract

“…To identify the healthiness and disease of the plant, the machine learning technique [3]- [5] was often applied to gain accuracy and performance. The clustering algorithm, as defined by L.a.b (value from the CIELAB colour scale) and coordinates of x and y-axis of the pixels [6], [7]. The input images are segmented into atomic regions.…”

Section: Related Workmentioning

confidence: 99%

Time-Series Classification Vegetables in Detecting Growth Rate Using Machine Learning

et al. 2021

View full text Add to dashboard Cite

IoT based innovative irrigation management systems can help in attaining optimum water-resource utilisation in the exactness farming landscape. This paper presents a clustering of unsupervised learning based innovative system to forecast the irrigation requirements of a field using the sensing of a ground parameter such as soil moisture, light intensity, temperature, and humidity. The entire system has been established and deployed. The sensor node data is gained through a serial monitor from Arduino IDE software collected directly and saved using the computer. Orange and MATLAB software is used to apply machine learning for the visualisation, and the decision support system delivers real-time information insights based on the analysis of sensors data. The plants organise either water or non-water includes weather conditions to gain various types of results. kNN reached 100.0%, SVM achieved 99.0% owhile Naïve Bayes achieved 87.40%.

show abstract

“…These efforts would give far-reaching consequences on UAV-WSN-IoT roles in smart farming and precision agriculture system construction ( Saif et al, 2017 ; Almalki et al, 2021 ). In addition, with the development of artificial intelligence, UAV-RSP can take advantage of high throughput to monitor the nutrient level, plant disease, and insect pest automatically ( Freitas et al, 2020 ; Tetila et al, 2020 ; Lpo et al, 2021 ). Therefore, UAV-RSP, coupled with IoT and 5G technology, plays a more and more important role in building the intelligent monitoring network for smart inspection of plant growth status ( Syed et al, 2020 ; Almalki et al, 2021 ).…”

Section: Introductionmentioning

confidence: 99%

Identification of High Nitrogen Use Efficiency Phenotype in Rice (Oryza sativa L.) Through Entire Growth Duration by Unmanned Aerial Vehicle Multispectral Imagery

et al. 2021

View full text Add to dashboard Cite

Identification of high Nitrogen Use Efficiency (NUE) phenotypes has been a long-standing challenge in breeding rice and sustainable agriculture to reduce the costs of nitrogen (N) fertilizers. There are two main challenges: (1) high NUE genetic sources are biologically scarce and (2) on the technical side, few easy, non-destructive, and reliable methodologies are available to evaluate plant N variations through the entire growth duration (GD). To overcome the challenges, we captured a unique higher NUE phenotype in rice as a dynamic time-series N variation curve through the entire GD analysis by canopy reflectance data collected by Unmanned Aerial Vehicle Remote Sensing Platform (UAV-RSP) for the first time. LY9348 was a high NUE rice variety with high Nitrogen Uptake Efficiency (NUpE) and high Nitrogen Utilization Efficiency (NUtE) shown in nitrogen dosage field analysis. Its canopy nitrogen content (CNC) was analyzed by the high-throughput UAV-RSP to screen two mixed categories (51 versus 42 varieties) selected from representative higher NUE indica rice collections. Five Vegetation Indices (VIs) were compared, and the Normalized Difference Red Edge Index (NDRE) showed the highest correlation with CNC (r = 0.80). Six key developmental stages of rice varieties were compared from transplantation to maturation, and the high NUE phenotype of LY9348 was shown as a dynamic N accumulation curve, where it was moderately high during the vegetative developmental stages but considerably higher in the reproductive developmental stages with a slower reduction rate. CNC curves of different rice varieties were analyzed to construct two non-linear regression models between N% or N% × leaf area index (LAI) with NDRE separately. Both models could determine the specific phenotype with the coefficient of determination (R2) above 0.61 (Model I) and 0.86 (Model II). Parameters influencing the correlation accuracy between NDRE and N% were found to be better by removing the tillering stage data, separating the short and long GD varieties for the analysis and adding canopy structures, such as LAI, into consideration. The high NUE phenotype of LY9348 could be traced and reidentified across different years, locations, and genetic germplasm groups. Therefore, an effective and reliable high-throughput method was proposed for assisting the selection of the high NUE breeding phenotype.

show abstract

Detection and classification of soybean pests using deep learning with UAV images

Cited by 169 publications

References 44 publications

A Novel Technique for Classifying Bird Damage to Rapeseed Plants Based on a Deep Learning Algorithm

A Novel Technique for Classifying Bird Damage to Rapeseed Plants Based on a Deep Learning Algorithm

Time-Series Classification Vegetables in Detecting Growth Rate Using Machine Learning

Identification of High Nitrogen Use Efficiency Phenotype in Rice (Oryza sativa L.) Through Entire Growth Duration by Unmanned Aerial Vehicle Multispectral Imagery

Contact Info

Product

Resources

About