Combining news media and AI to rapidly identify flooded buildings
2021-04-16
(Press-News.org) Artificial intelligence (AI) has sped up the process of detecting flooded buildings immediately after a large-scale flood, allowing emergency personnel to direct their efforts efficiently. Now, a research group from Tohoku University has created a machine learning (ML) model that uses news media photos to identify flooded buildings accurately within 24 hours of the disaster.
Their research was published in the journal Remote Sensing on April 5, 2021.
"Our model demonstrates how the rapid reporting of news media can speed up and increase the accuracy of damage mapping activities, accelerating disaster relief and response decisions, said Shunichi Koshimura of Tohoku University's International Research Institute of Disaster Science and co-author of the study.
ML and deep learning algorithms are tailored to classify objects through image analysis. For AI and ML to be effective, data is needed to train the model - flood data in the current case.
Although flood data can be collected from previous events, it will inadvertently lead to problems on account of every event being different and subject to the local characteristics of the flooded area. Thus, onsite information has higher reliability.
News crews and media teams are often the first on the scene of a disaster to broadcast images to viewers at home, and the research team recognized that this information too could be used in AI algorithms.
They applied their model to Mabi-cho, Kurashiki city in Okayama Prefecture, which was affected by the heavy rains across western Japan in 2018.
First, researchers identified press photos and geolocated them based on landmarks and other clues appearing in the photo. Next, they used synthetic aperture radar (SAR) PALSAR-2 images provided by JAXA to discretize flooded and non-flooded conditions of unknown areas.
Here, SAR images can be employed to classify water bodies since microwaves irradiate differently on wet and dry surfaces. A support vector machine (SVM), one of the machine learning techniques, was used to classify buildings surrounded by floodwaters or within non-flooded areas.
"The performance of our model resulted in an 80% estimation accuracy," added Koshimura.
Looking ahead, the research group will explore the applicability of news media databases from past events as training datasets for developing AI Models at present situations to increase the accuracy and speed of classification.
INFORMATION:
[Attachments] See images for this press release:
ELSE PRESS RELEASES FROM THIS DATE:
2021-04-16
Collaborative research of the University of Jyvaskyla and Natural Research Institute Finland presents new evidence of the effects of enriched rearing on well-being of aquaculture fishes. The research demonstrates that stone enrichments that have been previously conditioned in lake water significantly improve survival of fish compared to clean stones. Also a higher number of stones has a similar positive effect. The results have practical implications for prevention of aquaculture diseases. The study was published in Antibiotics in March 2021.
The volume of aquaculture is continuously increasing. Parasitic diseases represent a significant threat to farmed fishes and ecological solutions to minimize use of medication are being sought.
Enriched rearing, where rearing tanks ...
2021-04-16
A survey of star formation activity in the Orion Nebula Cluster found similar mass distributions for newborn stars and dense gas cores, which may evolve into stars. Counterintuitively, this means that the amount of gas a core accretes as it develops, and not the initial mass of the core, is the key factor in deciding the final mass of the produced star.
The Universe is populated with stars of various masses. Dense cores in clouds of interstellar gas collapse under their own gravity to form stars, but what determines the final mass of the star remains an ...
2021-04-16
Academic medical centers continuously strive to enhance patient care. One of the major mechanisms to improve patient health outcomes is through translational research - bringing research breakthroughs from the lab to patients via clinical trials. Making clinical trials more efficient, and ultimately more successful, would significantly advance patient care. However, fragmentation of the relevant data necessary to implement improvements to translational science is a significant barrier.
While some bioinformatic tools have attempted to address this problem, they often lacked the ability to assess the efficiency of translational science. ...
2021-04-16
Gas accidents such as toxic gas leakage in factories, carbon monoxide leakage of boilers, or toxic gas suffocation during manhole cleaning continue to claim lives and cause injuries. Developing a sensor that can quickly detect toxic gases or biochemicals is still an important issue in public health, environmental monitoring, and military sectors. Recently, a research team at POSTECH has developed an inexpensive, ultra-compact wearable hologram sensor that immediately notifies the user of volatile gas detection.
A joint research team led by Professor Junsuk Rho of departments of mechanical and chemical engineering and Dr. Inki Kim of Department of Mechanical ...
2021-04-16
A person who owns a car or who has a college education may be less vulnerable to COVID-19, according to an analysis of cases in Tehran, Iran, one of the early epicenters of the pandemic. While such variables do not inherently lower a person's risk, they do indicate an infrastructure of protection that persists despite how densely populated a person's district might be.
The international collaboration published their results on April 3 in Sustainable Cities and Society.
"In the past few decades, there have been various efforts aimed at increasing urban density to enhance efficiency and contribute to climate change mitigation -- but the COVID-19 pandemic has brought questions about ...
2021-04-16
The nervous system is the internet of the human body and can in the same way transfer signals over long distances very quickly. Some of the most important elements in this signaling are the axons. They are projections of the nerve cells which send signals to other nerve cells or muscles. For instance, axons that jut out from nerve cells in the spinal cord can be over one meter long.
Researchers from the Faculty of Health and Medical Sciences at the University of Copenhagen have now in a new study examined how signal molecules are transported in ...
2021-04-16
Scientists have called for the use of climate projections in conservation planning, to ensure that areas most at risk from biodiversity loss and climate impacts are protected. Protected areas are often created in areas of low population density and remote locations, rather than because of their biodiversity conservation potential. Conservation planning in tropical forests especially tends to be less rigorous and climate rarely taken into account, they said.
But climate projections and science-based information are critical to actively seek out and safeguard areas where species are most at risk and ecosystems are fast declining, said authors of the ...
2021-04-16
New research has shed light on how autism-spectrum disorder (ASD) manifests in the brains of girls, prompting the scientists to warn that conclusions drawn from studies conducted primarily in boys should not be assumed to hold true for girls.
The researchers discovered that there is a significant difference in the genes and "genetic burden" that underpin the condition in girls and boys. They also identified specific ways the brains of girls with ASD respond differently to social cues such as facial expressions and gestures than do those of girls without ASD.
"This new study provides us with a roadmap for understanding how to better match current and future evidenced-based interventions to underlying ...
2021-04-16
A candidate vaccine that could provide protection against the COVID-19 virus and other coronaviruses has shown promising results in early animal testing.
The candidate coronavirus vaccines, created by Virginia Tech's University Distinguished Professor X.J. Meng and UVA Health's Professor Steven L. Zeichner, prevented pigs from being becoming ill with a pig coronavirus, porcine epidemic diarrhea virus (PEDV).
The researchers have recently published their findings in the Proceedings of the National Academy of Sciences.
"The candidate vaccine was developed using an innovative vaccine platform targeting a highly conserved genomic region of coronaviruses," said Meng, a University ...
2021-04-16
Angiogenesis is a process of new vessel formation that is activated both in physiological (tissue repair, reproduction, etc.) and pathological (myocardial infarction, diabetic retinopathy, cancer, etc.) conditions. The process is carried out by endothelial cells and includes their proliferation, migration and arrangement in tubes. Angiogenesis regulation is precise and is mainly mediated by pro-angiogenic factors, such as vascular endothelial growth factor (VEGF), which in turn promote different signalling pathways leading to an increase of intracellular Ca2+ concentrations.
The researchers from the Cardiovascular Pathophysiology group at the Institute ...
LAST 30 PRESS RELEASES:
[Press-News.org] Combining news media and AI to rapidly identify flooded buildings
