Field rice panicle detection and counting based on deep learning

Wang, Xinyi; Yang, Wanneng; Lv, Qiucheng; Huang, Chenglong; Liang, Xinlian; Chen, Guoxing; Xiong, Lizhong; Duan, Lingfeng

doi:10.3389/fpls.2022.966495

Cited by 21 publications

(21 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, while previous works mainly relied on larger backbone networks or higher-resolution input images to achieve higher accuracy, balancing all dimensions of network width/depth/resolution is crucial when considering both accuracy and efficiency (Tan and Le, 2019;Dollaŕ et al, 2021;Tan and Le, 2021;Wang et al, 2022). Recent studies have shown that carefully designed lightweight networks can achieve comparable performance to their heavy counterparts with much less computational cost (Howard et al, 2017;Tan et al, 2020).…”

Section: Discussionmentioning

confidence: 99%

“…The field is mainly divided into two-stage algorithms represented by Faster Region-based Convolutional Neural Network (Faster RCNN) (Chen and Gupta, 2017) and one-stage algorithms represented by You Only Look Once (YOLOv3) (Redmon and Farhadi, 2018). Among them, CenterNet (Duan et al, 2019), EfficientDet (Tan et al, 2020), RetinaNet (Lin et al, 2017), Yolov7-tiny (Wang et al, 2022) have achieved SOTA performance in many different fields.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

TasselLFANet: a novel lightweight multi-branch feature aggregation neural network for high-throughput image-based maize tassels detection and counting

et al. 2023

Front. Plant Sci.

View full text Add to dashboard Cite

Accurately and rapidly counting the number of maize tassels is critical for maize breeding, management, and monitoring the growth stage of maize plants. With the advent of high-throughput phenotyping platforms and the availability of large-scale datasets, there is a pressing need to automate this task for genotype and phenotype analysis. Computer vision technology has been increasingly applied in plant science, offering a promising solution for automated monitoring of a large number of plants. However, the current state-of-the-art image algorithms are hindered by hardware limitations, which compromise the balance between algorithmic capacity, running speed, and overall performance, making it difficult to apply them in real-time sensing field environments. Thus, we propose a novel lightweight neural network, named TasselLFANet, with an efficient and powerful structure for accurately and efficiently detecting and counting maize tassels in high spatiotemporal image sequences. Our proposed approach improves the feature-learning ability of TasselLFANet by adopting a cross-stage fusion strategy that balances the variability of different layers. Additionally, TasselLFANet utilizes multiple receptive fields to capture diverse feature representations, and incorporates an innovative visual channel attention module to detect and capture features more flexibly and precisely. We conducted a series of comparative experiments on a new, highly informative dataset called MrMT, which demonstrate that TasselLFANet outperforms the latest batch of lightweight networks in terms of performance, flexibility, and adaptability, achieving an F1 measure value of 94.4%, a mAP.@5 value of 96.8%, and having only 6.0M parameters. Moreover, compared with the regression-based TasselNetV3-Seg† model, our proposed model achieves superior counting performance, with a mean absolute error (MAE) of 1.80, a root mean square error (RMSE) of 2.68, and a R2 of 0.99. The proposed model meets the accuracy and speed requirements of the vision system in maize tassel detection. Furthermore, our proposed method is reliable and unaffected by geographical changes, providing essential technical support for computerized counting in the field.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

TasselLFANet: a novel lightweight multi-branch feature aggregation neural network for high-throughput image-based maize tassels detection and counting

et al. 2023

Front. Plant Sci.

View full text Add to dashboard Cite

show abstract

“…However, there were difficulties in actual field testing for mobile applications. Wang et al (2022b) proposed a new method to remove repeated detections and achieved an accuracy of 92.77%. The methods, however, need to be optimized for UAV images and different density identification.…”

Section: Introductionmentioning

confidence: 99%

“…However, there were difficulties in actual field testing for mobile applications. Wang et al. (2022b) proposed a new method to remove repeated detections and achieved an accuracy of 92.77%.…”

Section: Introductionmentioning

confidence: 99%

Universal detection of curved rice panicles in complex environments using aerial images and improved YOLOv4 model

et al. 2022

View full text Add to dashboard Cite

Accurate and rapid identification of the effective number of panicles per unit area is crucial for the assessment of rice yield. As part of agricultural development, manual observation of effective panicles in the paddy field is being replaced by unmanned aerial vehicle (UAV) imaging combined with target detection modeling. However, UAV images of panicles of curved hybrid Indica rice in complex field environments are characterized by overlapping, blocking, and dense distribution, imposing challenges on rice panicle detection models. This paper proposes a universal curved panicle detection method by combining UAV images of different types of hybrid Indica rice panicles (leaf-above-spike, spike-above-leaf, and middle type) from four ecological sites using an improved You Only Look Once version 4 (YOLOv4) model. MobileNetv2 is used as the backbone feature extraction network based on a lightweight model in addition to a focal loss and convolutional block attention module for improved detection of curved rice panicles of different varieties. Moreover, soft non-maximum suppression is used to address rice panicle occlusion in the dataset. This model yields a single image detection rate of 44.46 FPS, and mean average precision, recall, and F1 values of 90.32%, 82.36%, and 0.89%, respectively. This represents an increase of 6.2%, 0.12%, and 16.24% from those of the original YOLOv4 model, respectively. The model exhibits superior performance in identifying different strain types in mixed and independent datasets, indicating its feasibility as a general model for detection of different types of rice panicles in the heading stage.

show abstract

“…At present, convolutional neural networks (CNN) are increasingly and widely used for plant and animal image detection ( Liu and Wang, 2020 ; Peng and Wang, 2022 ; Wang et al, 2022 ). The secondary detector represented by R-CNN for fast extraction and learning of image features was applied to the automated detection of batch images.…”

Section: Introductionmentioning

confidence: 99%

WSVAS: A YOLOv4 -based phenotyping platform for automatically detecting the salt tolerance of wheat based on seed germination vigour

Han

Liu

et al. 2022

Front. Plant Sci.

View full text Add to dashboard Cite

Salt stress is one of the major environmental stress factors that affect and limit wheat production worldwide. Therefore, properly evaluating wheat genotypes during the germination stage could be one of the effective ways to improve yield. Currently, phenotypic identification platforms are widely used in the seed breeding process, which can improve the speed of detection compared with traditional methods. We developed the Wheat Seed Vigour Assessment System (WSVAS), which enables rapid and accurate detection of wheat seed germination using the lightweight convolutional neural network YOLOv4. The WSVAS system can automatically acquire, process and analyse image data of wheat varieties to evaluate the response of wheat seeds to salt stress under controlled environments. The WSVAS image acquisition system was set up to continuously acquire images of seeds of four wheat varieties under three types of salt stress. In this paper, we verified the accuracy of WSVAS by comparing manual scoring. The cumulative germination curves of wheat seeds of four genotypes under three salt stresses were also investigated. In this study, we compared three models, VGG16 + Faster R-CNN, ResNet50 + Faster R-CNN and YOLOv4. We found that YOLOv4 was the best model for wheat seed germination target detection, and the results showed that the model achieved an average detection accuracy (mAP) of 97.59%, a recall rate (Recall) of 97.35% and the detection speed was up to 6.82 FPS. This proved that the model could effectively detect the number of germinating seeds in wheat. In addition, the germination rate and germination index of the two indicators were highly correlated with germination vigour, indicating significant differences in salt tolerance amongst wheat varieties. WSVAS can quantify plant stress caused by salt stress and provides a powerful tool for salt-tolerant wheat breeding.

show abstract

Field rice panicle detection and counting based on deep learning

Cited by 21 publications

References 35 publications

TasselLFANet: a novel lightweight multi-branch feature aggregation neural network for high-throughput image-based maize tassels detection and counting

TasselLFANet: a novel lightweight multi-branch feature aggregation neural network for high-throughput image-based maize tassels detection and counting

Universal detection of curved rice panicles in complex environments using aerial images and improved YOLOv4 model

WSVAS: A YOLOv4 -based phenotyping platform for automatically detecting the salt tolerance of wheat based on seed germination vigour

Contact Info

Product

Resources

About