Panicle Counting in UAV Images For Estimating Flowering Time in Sorghum

Cai, Enyu; Baireddy, Sriram; Yang, Clifford; Crawford, Melba M.; Delp, Edward J.

doi:10.48550/arxiv.2107.07308

Cited by 3 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine learning algorithms have been applied to tasks such as panicle detection in sorghum [4] and tassel detection in maize [22,23]. To generate tassel count estimates for each of our plot-level images and to then track how the counts for each individual plot change as the plants develop to maturity, we investigated the accuracy of both a CBR (TasselNet) and a CBD approach (see Section 2), following the procedure as illustrated with the R-CNN used for the CBD approach (Figure 2).…”

Section: Filtered Image Feeds Into a Cbr Algorithm Implemented In Tas...mentioning

confidence: 99%

“…Sensors 2024, 24, 2172 2 of 14 are attractive, as they enable detailed quantification of traits related to flowering time, i.e., panicle counts in sorghum [4] and the duration of flowering time for a given genotype. The latter trait will contribute to making more valuable crossing decisions from pollen donor to pollen receiver, as some of the female and male organs might not be ready at the same time.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Image Filtering to Improve Maize Tassel Detection Accuracy Using Machine Learning Algorithms

Rodene,

Fernando,

Piyush

et al. 2024

Sensors

View full text Add to dashboard Cite

Unmanned aerial vehicle (UAV)-based imagery has become widely used to collect time-series agronomic data, which are then incorporated into plant breeding programs to enhance crop improvements. To make efficient analysis possible, in this study, by leveraging an aerial photography dataset for a field trial of 233 different inbred lines from the maize diversity panel, we developed machine learning methods for obtaining automated tassel counts at the plot level. We employed both an object-based counting-by-detection (CBD) approach and a density-based counting-by-regression (CBR) approach. Using an image segmentation method that removes most of the pixels not associated with the plant tassels, the results showed a dramatic improvement in the accuracy of object-based (CBD) detection, with the cross-validation prediction accuracy (r2) peaking at 0.7033 on a detector trained with images with a filter threshold of 90. The CBR approach showed the greatest accuracy when using unfiltered images, with a mean absolute error (MAE) of 7.99. However, when using bootstrapping, images filtered at a threshold of 90 showed a slightly better MAE (8.65) than the unfiltered images (8.90). These methods will allow for accurate estimates of flowering-related traits and help to make breeding decisions for crop improvement.

show abstract

Section: Filtered Image Feeds Into a Cbr Algorithm Implemented In Tas...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Image Filtering to Improve Maize Tassel Detection Accuracy Using Machine Learning Algorithms

Rodene,

Fernando,

Piyush

et al. 2024

Sensors

View full text Add to dashboard Cite

show abstract

“…1. revisit breeding objectives to consider (1) improving for insect and disease resistance such as sugarcane aphid and ALS herbicide tolerance, and (2) expanding market uses, including biofuels, and ecosystem services for double cropping such as sorghum/potato in the southeast United States, 2. improved field-based phenotyping through the deployment of digital technologies to image panicles (Ghosal et al, 2019;Gonzalo-Martin et al, 2021;Lin & Guo, 2020), flowering time (Cai et al, 2021), biomass (Chung et al, 2017), architecture (McCormick et al, 2016, light use efficiency (Furbank et al, 2019;Wu et al, 2019), staygreen traits (Barnhart et al, 2021), and root architecture (Singh et al, 2011), 3. use of managed stress environments and enviromics to select for both yield stability and yield potential under top-end as well as dryland stressed environments (Carcedo et al, 2022;Cooper & Messina, 2021;, 4. implementation of crop growth model-whole-genome predictions (Messina et al, 2018;Diepenbrock et al, 2022) to focus on drought-adapted traits and gap analyses methodology (Cooper et al, 2020), 5. development of management technology and crop protection tools (Ciampitti et al, 2019) for on-farm…”

Section: Perspectivesmentioning

confidence: 99%

Retrospective study in US commercial sorghum breeding: I. Genetic gain in relation to relative maturity

et al. 2023

View full text Add to dashboard Cite

Sorghum [Sorghum bicolor (L.) Moench] is an important staple food for human consumption and a source of animal feed in the semiarid regions of the world. Sustained positive rates of crop improvement are necessary to supply food and feed to a growing population. However, land allocated to sorghum and its inclusion in production systems has been in constant decline. Here we report the rate of sorghum genetic gain in a commercial breeding program in the United States and provide evidence that a modest yield improvement is an important factor limiting land allocation to this crop. A 6‐year study that evaluated 50 sorghum genotypes commercialized between the decades of 1960 and 2010 was conducted in 19 environments within the US Sorghum Belt region. Yield varied between 500 and 850 g m−2. Here we show a positive rate of genetic gain of 2.63 g m−2 y−1 on average across three different maturity groups grown in the United States. Rates ranged from 2.1 to 4.3 g m−2 y−1 across maturity groups. This result contrasts with a stagnant rate of crop improvement for many regions of the world, yet the rates are insufficient to reverse the negative trend in planted area. Breeding technologies are proposed to hasten genetic gain in sorghum to reverse the loss of on‐farm agricultural biodiversity.

show abstract

“…Once the images are filtered by the ExG, the second part of the filtering process is to remove the foliage, leaving behind only the tassel pixels. This is accomplished by using a second formula: (2) [2(𝐺 * + 𝑅 * )] 2 − 𝐵 * As described with the ExG above, the resulting value is then rescaled to range from 0 to 255. This formula was selected out of a variety of other possibilities by in-house tests which indicated it had the best performance with distinguishing between tassels and foliage.…”

Section: Procedures For Image Filtrationmentioning

confidence: 99%

“…Conventional methods to manually measure these traits are both tedious and time-consuming. Automated methods of both detecting and counting tassels or panicles would allow traits related to flowering time to be quantified in detail, i.e., panicle counting in sorghum [2]. Recently, several machine learning algorithms, such as ResNet [3], YOLO [4], and Faster R-CNN [5], have been developed for tassel detection from unmanned aerial vehicle (UAV)-based photography [6,7,8,9,10].…”

Section: Introductionmentioning

confidence: 99%

Machine learning-based tassel detection for time-series high throughput plant phenotyping

Rodene

Schnable

et al. 2022

Preprint

View full text Add to dashboard Cite

Unmanned aerial vehicle (UAV)-based imagery has become widely used in collecting agronomic traits, enabling a much greater volume of data to be generated in a time-series manner. As one of the cuttingedge imagery analysis tools, machine learning-based object detection provides automated techniques to analyze these imagery data. In our previous study, UAVs have been used to collect aerial photography for field trials of 233 diverse inbred lines, grown under different nitrogen treatments. Images were collected during different plant developmental stages throughout the growing season. This dataset of images has here been used in developing machine learning techniques to obtain automated tassel counts at the plot level through the season. To improve detection accuracy, we have developed an image segmentation method to remove non-tassel pixels and then feed these filtered images into machine learning algorithms. As a result, our method showed a significant improvement in the accuracy of maize tassel detection. This method can be used in future research to produce time-series counts of tassels at the plot level, and will allow for accurate estimates of flowering-related traits, such as the earliest detected flowering date and the duration of each plot's flowering period. This phenotypic data and the traitassociated genes provide new opportunities for crop improvement and to facilitate future plant breeding.

show abstract

Panicle Counting in UAV Images For Estimating Flowering Time in Sorghum

Cited by 3 publications

References 17 publications

Image Filtering to Improve Maize Tassel Detection Accuracy Using Machine Learning Algorithms

Image Filtering to Improve Maize Tassel Detection Accuracy Using Machine Learning Algorithms

Retrospective study in US commercial sorghum breeding: I. Genetic gain in relation to relative maturity

Machine learning-based tassel detection for time-series high throughput plant phenotyping

Contact Info

Product

Resources

About