Dynamic detection method for falling ears of maize harvester based on improved YOLO-V4

Gao, Ang; Geng, Aijun; Ji, Zhang; Hu, Xiaolong; Li, Ké

doi:10.25165/j.ijabe.20221503.6660

Cited by 3 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The young citrus fruit targets in the images were labelled using Labelimg image annotation tool and the images and labelled data were stored in PASCAL VOC format. After completing the labelling, the dataset was divided into training, testing and validation sets in the ratio of 7:2:1 ( Gao et al., 2022 ).…”

Section: Experiments and Methodsmentioning

confidence: 99%

Fruits hidden by green: an improved YOLOV8n for detection of young citrus in lush citrus trees

Ang,

Zhiwei,

Wei

et al. 2024

Front. Plant Sci.

View full text Add to dashboard Cite

In order to address the challenges of inefficiency and insufficient accuracy in the manual identification of young citrus fruits during thinning processes, this study proposes a detection methodology using the you only look once for complex backgrounds of young citrus fruits (YCCB-YOLO) approach. The method first constructs a dataset containing images of young citrus fruits in a real orchard environment. To improve the detection accuracy while maintaining the computational efficiency, the study reconstructs the detection head and backbone network using pointwise convolution (PWonv) lightweight network, which reduces the complexity of the model without affecting the performance. In addition, the ability of the model to accurately detect young citrus fruits in complex backgrounds is enhanced by integrating the fusion attention mechanism. Meanwhile, the simplified spatial pyramid pooling fast-large kernel separated attention (SimSPPF-LSKA) feature pyramid was introduced to further enhance the multi-feature extraction capability of the model. Finally, the Adam optimization function was used to strengthen the nonlinear representation and feature extraction ability of the model. The experimental results show that the model achieves 91.79% precision (P), 92.75% recall (R), and 97.32% mean average precision (mAP)on the test set, which were improved by 1.33%, 2.24%, and 1.73%, respectively, compared with the original model, and the size of the model is only 5.4 MB. This study could meet the performance requirements for citrus fruit identification, which provides technical support for fruit thinning.

show abstract

Section: Experiments and Methodsmentioning

confidence: 99%

Fruits hidden by green: an improved YOLOV8n for detection of young citrus in lush citrus trees

Ang,

Zhiwei,

Wei

et al. 2024

Front. Plant Sci.

View full text Add to dashboard Cite

show abstract

“…CNN neural networks are widely used in precision agriculture and smart agriculture, such as automatic species identification [10], disease identification [11,12] and fruit ripeness analysis [13][14][15]. Most of the research on crop and weed localization is based on object detection networks, such as You Only Look Once (YOLO) [16][17][18][19] series models and Region-CNN (RCNN) series models [20][21][22]. Zou et al [23] combined images with and without weeds to generate new weed images, and trained a semantic segmentation network called UNet, obtaining an accuracy of 92.21%.…”

Section: Introductionmentioning

confidence: 99%

Real-Time Detection of Crops with Dense Planting Using Deep Learning at Seedling Stage

Kong

Zhai

et al. 2023

Agronomy

View full text Add to dashboard Cite

Crop seedlings are similar in appearance to weeds, making crop detection extremely difficult. To solve the problem of detecting crop seedlings in complex field environments, a seedling dataset with four crops was constructed in this study. The single leaf labeling method was proposed as an alternative to conventional labeling approaches to improve the detection accuracy for dense planting crops. Second, a seedling detection network based on YOLOv5 and a transformer mechanism was proposed, and the effects of three features (query, key and value) in the transformer mechanism on the detection accuracy were explored in detail. Finally, the seedling detection network was optimized into a lightweight network. The experimental results show that application of the single leaf labeling method could improve the mAP0.5 of the model by 1.2% and effectively solve the problem of missed detection. By adding the transformer mechanism module, the mAP0.5 was improved by 1.5%, enhancing the detection capability of the model for dense and obscured targets. In the end, this study found that query features had the least impact on the transformer mechanism, and the optimized model improved the computation speed by 23 ms·frame−1 on the intelligent computing platform Jetson TX2, providing a theoretical basis and technical support for real-time seedling management.

show abstract

“…Zhou Liping et al proposed an improved Otsu algorithm based ear root feature region detection for pig thermal infrared images [10] , which cannot accurately detect the ear root region of pig heads under dynamic conditions. Liu Gang et al proposed a thermal infrared video detection method for pig ear root temperature based on improved YOLO v4 [11][12] , which removes skewed frames of pig posture while retaining upright frames, which can easily cause information loss. Zhao Haitao proposed a pig body temperature detection and key temperature measurement part recognition based on infrared thermal imaging technology.…”

Section: Introductionmentioning

confidence: 99%

Detection of ear root feature region in heterogeneous images based on pig head posture

2023

AJCIS

View full text Add to dashboard Cite

In the scene of pig thermal infrared temperature measurement, the temperature of the ear root is closest to the body temperature of the pig. In order to address the problem of inaccurate positioning of the ear root area in the thermal infrared image of the pig under motion due to changes in the pig's head posture and single pixel information, based on the research of automatic inspection and temperature measurement robots, it is proposed to establish an ear root area detection model through visible light images, and map the detection results of the model to the thermal infrared image with the same contour information after registration, achieve accurate localization of the pig ear root area in the thermal infrared image. This algorithm achieves the detection of key points in pig heads in visible light images by improving the YOLOv8 network. By analyzing the position information of key points in pig heads through light shadow projection and dynamic coordinate system adjustment methods, the pig head posture and adjustment angle are obtained. The visible light image of the pig's head after angle adjustment is redivided into ear root regions to correct the impact of pig movement on ear root region division. By using AKAZE and GMA algorithms to extract and register features from heterogeneous images, visible and thermal infrared images with the same contour and scale information are obtained. Finally, the detection results of the ear root region in the visible light image are mapped to the thermal infrared image. Through testing on 2650 datasets, the accuracy of key point and category detection was 99%, the number of key point detection model parameters decreased by 41%, and the registration error was only 1.6%. The accuracy of ear root localization in heterogeneous images reached over 99%, making it easy to achieve accurate temperature measurement.

show abstract

Dynamic detection method for falling ears of maize harvester based on improved YOLO-V4

Cited by 3 publications

References 25 publications

Fruits hidden by green: an improved YOLOV8n for detection of young citrus in lush citrus trees

Fruits hidden by green: an improved YOLOV8n for detection of young citrus in lush citrus trees

Real-Time Detection of Crops with Dense Planting Using Deep Learning at Seedling Stage

Detection of ear root feature region in heterogeneous images based on pig head posture

Contact Info

Product

Resources

About