(Press-News.org)
To improve grape yield predictions, automated berry counting has emerged as a crucial yet challenging task due to the dense distribution and occlusion of berries. While grape cultivation is a significant global economic activity, traditional manual counting methods are inaccurate and inefficient. Recent research has shifted towards deep learning and computer vision, employing detection and density estimation techniques for more precise counts. However, these methods grapple with the variability of farmland and high occlusion rates, leading to significant counting errors. Additionally, creating high-performance algorithms demands expensive data labeling, making it a significant hurdle for widespread use and adaptability in the field. The need for a cost-effective, accurate automated counting method remains a pressing research problem.
In November 2023, Plant Phenomics published a research article entitled by “Semi-supervised Counting of Grape Berries in the Field Based on Density Mutual Exclusion”.
To effectively and cost-efficiently count grape berries, the research presents CDMENet, a semi-supervised method that uses VGG16 for image feature extraction and density mutual exclusion to understand spatial patterns and unlabeled data. The method also employs a density difference loss to amplify feature differences between varying density levels. Conducted on an Ubuntu system with Python and deep learning frameworks, the algorithm's performance was tested using a specified hardware setup, preprocessing techniques, and training details such as learning rates and optimization methods. The results were promising, with CDMENet outperformed both fully and semi-supervised counterparts in terms of Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and the coefficient of determination (R2). Particularly notable was its performance with limited labeled data, showcasing superior accuracy and reduced errors compared to other models. Additionally, the method's robustness was further evidenced in various ablation studies, demonstrating the significant roles of unlabeled data, density difference loss, and the number of auxiliary task predictors. Despite minor influences from prediction confidence thresholds, CDMENet maintained stable and robust performance.
In conclusion, CDMENet presents a viable solution for grape berry counting in fields, reducing the need for extensive manual labeling while improving accuracy and reducing errors. Its efficient use of unlabeled data, enhanced feature representation through density difference loss, and overall robust performance against a backdrop of varying conditions underscore its potential as a cost-effective tool in agricultural yield estimation. Future work might explore optimizing loss functions further and deploying the algorithm in field robots or other practical applications.
###
References
Authors
Yanan Li1,2†, Yuling Tang1,2*†, Yifei Liu1,2, and Dingrun Zheng1,2
†These authors contributed equally to this work.
Affiliations
1School of Computer Science and Engineering, School of Artificial Intelligence, Wuhan Institute of Technology, Wuhan 430205, China.
2Hubei Key Laboratory of Intelligent Robot, Wuhan Institute of Technology, Wuhan 430073, China.
About Yanan Li
She is currently a Lecturer with the School of Computer Science and Engineering, Wuhan Institute of Technology, Wuhan. Her research interests include computer vision and machine learning, with particular emphasis on image segmentation, domain adaptation, and various computer vision applications in agriculture.
END
Advancements in whole-genome sequencing have revolutionized plant species characterization, providing a wealth of genotypic data for analysis. The combination of genomic selection and neural networks, especially deep learning and autoencoders, has emerged as a promising method for predicting complex traits from this data. Despite the success in applications like plant phenotyping, challenges remain in accurately translating visual information from images into measurable data for genomic studies.
In November 2023, Plant ...
Erin Overstreet has been selected as the new executive director of the USC Stevens Center for Innovation where she will oversee the university’s commercialization of USC-driven intellectual property.
Overstreet’s expertise and experience embody technology transfer and innovation across the academic, educational, and venture capital sectors; such experience is critical for bridging USC research to a broadened, national technology transfer ecosystem, said Ishwar Puri, senior vice president of the Office of Research and Innovation.
“The university has the utmost confidence in Dr. Overstreet’s ability to ...
In the field of Landscape Architecture, Topography aims to study the complex and ongoing changing relationship between humans and the land through continuously updated and iterative tools and media. It maintains a balance between abstract concepts and concrete perceptions, which can both drive the development of science and technology in this field and hold on to openness to artistic expression. Thus, topographical design may be an effective way to help facilitate refining landscape design methods.
The work entitled “Can Topography ...
Associate Professor Andrew Baker from QUT School of Biology and Environmental Science said antechinuses are carnivorous marsupials well-known for suicidal sex sessions where all males die after the 1 to 3 week breeding period.
“During the breeding season, male and females mate promiscuously in frenzied bouts lasting as long as 14 hours. Certain stress-induced death follows for all males as surging testosterone causes cortisol to flood uncontrolled through the body, reaching pathological levels,” Professor Baker said.
“The males drop dead, which provides an opportunity ...
Protecting the world’s oceans against accelerating damage from human activities could be cheaper and take up less space than previously thought, new research has found.
The University of Queensland’s Professor Anthony Richardson collaborated on the study, which looks to halt the rapid decline of marine biodiversity from expanding industrial activities in marine areas beyond national jurisdictions (ABNJ).
“This ‘blue acceleration’ as we call it, has seen a greater diversity of stakeholders interested in ABNJs, such as the high seas and the international seabed beyond exclusive economic zones,” ...
Frequently using more than three strategies to stay alert while driving could be a sign of excessive sleepiness due to obstructive sleep apnoea (OSA), according to a study published today (Thursday) in ERJ Open Research [1].
People with OSA often snore loudly, their breathing starts and stops during the night, and they may wake up several times. Around one in five people are estimated to have OSA but the majority of sufferers do not realise they have a problem. OSA causes excessive sleepiness and people with untreated OSA are at higher risk of collisions on the road.
Researchers say that asking people ...
Globally, non-point source pollution is an important source of water quality deterioration in rivers and lakes. A ditch-pond system, consisting of ditches and ponds, is considered to be similar to free-surface wetlands, linking pollution sources to the receiving water bodies. The ditch-pond system includes vegetation, microorganisms and sediment, which can slow down the flow velocity and promote the precipitation of particulate matter carried by running water. At the same time, ditch and pond systems reduces nitrogen and phosphorus concentrations, and those of other nutrients entering the downstream water by means of plant absorption, sediment adsorption and microbial degradation, ...
The Lancet Child & Adolescent Health: Paediatric care for non-White children is universally worse across the USA; policy reform urgently needed to address disparities
Two-paper Series identifies pervasive racial inequities in paediatric care in the USA, and outlines policies to address structural racism embedded in wider sectors of society that shape children’s health.
A review of recent evidence reveals widespread patterns of inequitable care across paediatric specialties, including neonatal care, emergency medicine, surgery, developmental disabilities, mental ...
Peer reviewed – observational study - humans
Genotyping technology detects Covid variants more quickly and cheaply than ever before – according to research from the University of East Anglia and the UK Health Security Agency.
A new study published today reveals that the technique detects new variants almost a week more quickly than traditional whole genome sequencing methods.
The research team say that genotyping allowed Covid variant information to be more rapidly detected and communicated to frontline health protection professionals at the height of the pandemic.
Importantly, it helped to implement ...
People with obsessive-compulsive disorder (OCD) may have an increased risk of death from both natural and unnatural causes than those without the disorder, finds a study from Sweden published by The BMJ today.
The researchers point out that many of the natural causes of death are preventable, suggesting that better surveillance, prevention, and early intervention strategies should be implemented to reduce the risk of fatal outcomes in people with OCD.
OCD is typically a long term psychiatric disorder affecting about 2% of the population. It is characterised ...
