Presentation
Time:
M.S. Thesis Defense: Geng (Frank) Bai
Date:
9:30 am –
11:30 am
Virtual Location:
Zoom
Target Audiences:
M.S. Thesis Defense: Geng (Frank) Bai
Thursday, July 25, 2024
9:30 AM CST
Zoom: https://ncsu.zoom.us/j/4232052632?omn=96421560976
Meeting ID: 423 205 2632
“Quantifying Maize Tassel Traits for UNL Plant Phenotyping Facility through a Light-Weight Deep Learning Model”
Enhancing maize breeding programs is crucial for increasing yield, stress resistance, and nutrient use efficiency, especially in Nebraska, a leading corn-producing state in the United States. Tassel traits are vital for understanding reproductive development and optimizing pollination success, which directly impacts yield. The UNL Field Plant Phenotyping Facility, known as the NU-Spidercam, is a cutting-edge research infrastructure at the University of Nebraska-Lincoln (UNL) designed to advance plant phenotyping in agricultural research. Equipped with high-resolution cameras and imaging sensors, the Spidercam system enables researchers to precisely analyze the physical traits of tassels. Although existing publications provide public datasets and open-source models, they are not yet fully equipped for effective tassel detection and localization in the target environment. This study presents the development of a customized TasselLFANet model, a lightweight deep-learning model designed for real-time quantification of maize tassel location and size for the Spidercam system. Utilizing frequent imaging and transfer learning techniques, the model successfully outputs tassel count and total area of bounding boxes with limited labeled images. The model was deployed to predict tassel traits for a field experiment with 23 genotypes and 2 water treatments. The model showed significant performance improvements when the number of labeled images increased from 100 to over 600. The model successfully tracked tassel development with high temporal resolution, comparable to canopy height and NDVI in detecting the VT stage. Additionally, the model shows promise in informing the earliest timing for relatively accurate yield prediction using other phenotypic traits. Correlation analysis indicated medium or weak correlations between tassel traits and other phenotypic characteristics. The correlation between tassel traits and yield indicates its potential contribution to yield prediction as a unique feature in machine learning algorithms. However, the model struggled to detect small and occluded tassels in modern genotypes, indicating areas for future improvement. Overall, this study demonstrates that effective phenotyping tools (tassel traits detection in this case) can be developed with limited labeling effort through public datasets, open-source models, and transfer learning techniques.
Committee:
Hongfeng Yu, Advisor
Ashok Samal
Yufeng Ge
Thursday, July 25, 2024
9:30 AM CST
Zoom: https://ncsu.zoom.us/j/4232052632?omn=96421560976
Meeting ID: 423 205 2632
“Quantifying Maize Tassel Traits for UNL Plant Phenotyping Facility through a Light-Weight Deep Learning Model”
Enhancing maize breeding programs is crucial for increasing yield, stress resistance, and nutrient use efficiency, especially in Nebraska, a leading corn-producing state in the United States. Tassel traits are vital for understanding reproductive development and optimizing pollination success, which directly impacts yield. The UNL Field Plant Phenotyping Facility, known as the NU-Spidercam, is a cutting-edge research infrastructure at the University of Nebraska-Lincoln (UNL) designed to advance plant phenotyping in agricultural research. Equipped with high-resolution cameras and imaging sensors, the Spidercam system enables researchers to precisely analyze the physical traits of tassels. Although existing publications provide public datasets and open-source models, they are not yet fully equipped for effective tassel detection and localization in the target environment. This study presents the development of a customized TasselLFANet model, a lightweight deep-learning model designed for real-time quantification of maize tassel location and size for the Spidercam system. Utilizing frequent imaging and transfer learning techniques, the model successfully outputs tassel count and total area of bounding boxes with limited labeled images. The model was deployed to predict tassel traits for a field experiment with 23 genotypes and 2 water treatments. The model showed significant performance improvements when the number of labeled images increased from 100 to over 600. The model successfully tracked tassel development with high temporal resolution, comparable to canopy height and NDVI in detecting the VT stage. Additionally, the model shows promise in informing the earliest timing for relatively accurate yield prediction using other phenotypic traits. Correlation analysis indicated medium or weak correlations between tassel traits and other phenotypic characteristics. The correlation between tassel traits and yield indicates its potential contribution to yield prediction as a unique feature in machine learning algorithms. However, the model struggled to detect small and occluded tassels in modern genotypes, indicating areas for future improvement. Overall, this study demonstrates that effective phenotyping tools (tassel traits detection in this case) can be developed with limited labeling effort through public datasets, open-source models, and transfer learning techniques.
Committee:
Hongfeng Yu, Advisor
Ashok Samal
Yufeng Ge