Detection and quantification of cotton trichomes by deep learning algorithm

Luo, Keyan; Yu, Jihong; Wen, Shuhan; Li, Yicheng; Rong, Junkang; Ding, Mingquan

doi:10.1016/j.compag.2023.107936

Cited by 9 publications

(4 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, because Detectron2 incorporates several widely employed deep learning models for object detection and instance segmentation, it possesses the potential for future compatibility with a broader range of agricultural and industrial production scenarios. These scenarios may include tasks like recognizing plant fructifications and identifying crop pests, extending its applicability beyond the sole measurement of rapeseed pod phenotype omics data [ 62 , 63 , 64 , 65 , 66 ]. By combining machine vision, we also determined the length, width, and two-dimensional image area of the rapeseed pods in the image using a single coin as a reference.…”

Section: Discussionmentioning

confidence: 99%

Segmentation and Phenotype Calculation of Rapeseed Pods Based on YOLO v8 and Mask R-Convolution Neural Networks

Wang,

Liu,

et al. 2023

Plants

Self Cite

View full text Add to dashboard Cite

Rapeseed is a significant oil crop, and the size and length of its pods affect its productivity. However, manually counting the number of rapeseed pods and measuring the length, width, and area of the pod takes time and effort, especially when there are hundreds of rapeseed resources to be assessed. This work created two state-of-the-art deep learning-based methods to identify rapeseed pods and related pod attributes, which are then implemented in rapeseed pots to improve the accuracy of the rapeseed yield estimate. One of these methods is YOLO v8, and the other is the two-stage model Mask R-CNN based on the framework Detectron2. The YOLO v8n model and the Mask R-CNN model with a Resnet101 backbone in Detectron2 both achieve precision rates exceeding 90%. The recognition results demonstrated that both models perform well when graphic images of rapeseed pods are segmented. In light of this, we developed a coin-based approach for estimating the size of rapeseed pods and tested it on a test dataset made up of nine different species of Brassica napus and one of Brassica campestris L. The correlation coefficients between manual measurement and machine vision measurement of length and width were calculated using statistical methods. The length regression coefficient of both methods was 0.991, and the width regression coefficient was 0.989. In conclusion, for the first time, we utilized deep learning techniques to identify the characteristics of rapeseed pods while concurrently establishing a dataset for rapeseed pods. Our suggested approaches were successful in segmenting and counting rapeseed pods precisely. Our approach offers breeders an effective strategy for digitally analyzing phenotypes and automating the identification and screening process, not only in rapeseed germplasm resources but also in leguminous plants, like soybeans that possess pods.

show abstract

Section: Discussionmentioning

confidence: 99%

Segmentation and Phenotype Calculation of Rapeseed Pods Based on YOLO v8 and Mask R-Convolution Neural Networks

Wang,

Liu,

et al. 2023

Plants

Self Cite

View full text Add to dashboard Cite

show abstract

“…The percentage increase in performance metrics when compared to the control model for each precision, recall, mAP@0.5, and mAP@[0.5:0.95] value is 2.1%; 4.4%; 5.8%; and 1.0%. The results of this increase show a significant impact on the addition of preprocessing and augmentation processes that match the characteristics of the dataset to increase the value of model performance (Luo et al, 2023). However, in the length of time of the training process, the model with the addition of preprocessing and augmentation has the longest time compared to the other combinations, and that is even 2 times longer than the control model.…”

Section: Adjustment Of Preprocessing and Augmentation Processmentioning

confidence: 97%

Effect of Preprocessing and Augmentation Process in Development of a Deep Learning Model for Fusarium Detection in Shallots

Purwansya,

Solahudin,

Supriyanto

2024

JTEP-L

View full text Add to dashboard Cite

As the demand for shallot increases, wide-scale cultivation area must be managed efficiently. However, shallot productivity decreases every year because of plant diseases. Fusarium disease has an intensity up to 60% and can affect yield losses up to 50%. This study was conducted to develop the fusarium disease detection system for shallot using deep learning model and analyze the effect of preprocessing and augmentation adjustment. This study used YOLOv5 deep learning algorithm consisting of the following stages: (1) dataset acquisition, (2) dataset annotation, (3) dataset preprocessing and augmentation, (4) dataset training and validation, and (5) model testing and evaluation. A total 9,664 annotated dataset was trained to YOLOv5m pre-trained weights. Based on testing and evaluation results, precision, recall, and mean average precision (mAP) metrics of the model without preprocessing and augmentation were 55.5%; 54%; and 48.3% respectively. Metric values of the model were increased to 57.6%; 58.4%; and 54.1% respectively with adjustment of preprocessing and augmentation combination process. Percentage increase in metrics when compared to the control model for each value of precision, recall, and mAP were 2.1%; 4.4%; and 5.8%. This shows a significant impact on the addition of preprocessing and augmentation processes that match the characteristics of the dataset to increase the value of model performance. Keywords: Augmentation, Deep learning, Fusarium, Shallot.

show abstract

“…The authors used different dataset stratification strategies (leaf-based splits, yearbased splits, and environmental-based splits). Luo et al (2023) developed a deep learning approach to detect and quantify trichomes on cotton leaves and stems. The trichomes on the stem edge and leaf edge were photographed using an Olympus szx10 stereoscope with ×12.5 magnification.…”

Section: Related Workmentioning

confidence: 99%

Image-based classification of wheat spikes by glume pubescence using convolutional neural networks

Artemenko,

Genaev,

Epifanov

et al. 2024

Front. Plant Sci.

View full text Add to dashboard Cite

IntroductionPubescence is an important phenotypic trait observed in both vegetative and generative plant organs. Pubescent plants demonstrate increased resistance to various environmental stresses such as drought, low temperatures, and pests. It serves as a significant morphological marker and aids in selecting stress-resistant cultivars, particularly in wheat. In wheat, pubescence is visible on leaves, leaf sheath, glumes and nodes. Regarding glumes, the presence of pubescence plays a pivotal role in its classification. It supplements other spike characteristics, aiding in distinguishing between different varieties within the wheat species. The determination of pubescence typically involves visual analysis by an expert. However, methods without the use of binocular loupe tend to be subjective, while employing additional equipment is labor-intensive. This paper proposes an integrated approach to determine glume pubescence presence in spike images captured under laboratory conditions using a digital camera and convolutional neural networks.MethodsInitially, image segmentation is conducted to extract the contour of the spike body, followed by cropping of the spike images to an equal size. These images are then classified based on glume pubescence (pubescent/glabrous) using various convolutional neural network architectures (Resnet-18, EfficientNet-B0, and EfficientNet-B1). The networks were trained and tested on a dataset comprising 9,719 spike images.ResultsFor segmentation, the U-Net model with EfficientNet-B1 encoder was chosen, achieving the segmentation accuracy IoU = 0.947 for the spike body and 0.777 for awns. The classification model for glume pubescence with the highest performance utilized the EfficientNet-B1 architecture. On the test sample, the model exhibited prediction accuracy parameters of F1 = 0.85 and AUC = 0.96, while on the holdout sample it showed F1 = 0.84 and AUC = 0.89. Additionally, the study investigated the relationship between image scale, artificial distortions, and model prediction performance, revealing that higher magnification and smaller distortions yielded a more accurate prediction of glume pubescence.

show abstract

Detection and quantification of cotton trichomes by deep learning algorithm

Cited by 9 publications

References 24 publications

Segmentation and Phenotype Calculation of Rapeseed Pods Based on YOLO v8 and Mask R-Convolution Neural Networks

Segmentation and Phenotype Calculation of Rapeseed Pods Based on YOLO v8 and Mask R-Convolution Neural Networks

Effect of Preprocessing and Augmentation Process in Development of a Deep Learning Model for Fusarium Detection in Shallots

Image-based classification of wheat spikes by glume pubescence using convolutional neural networks

Contact Info

Product

Resources

About