news

jimu news reporter zhan ying

in the two schematic diagrams, 3,747 "black" craters and densely packed red dots showing damaged buildings record the tragic situation of the continuous bombing in the gaza strip.

on august 24, a paper titled "time-series satellite remote sensing reveals that war damage in the gaza strip continues to rise" was published online in the comprehensive international top journal national science review. it is understood that the paper is the latest research result of the team of academician li deren from the state key laboratory of surveying, mapping and remote sensing information engineering of wuhan university.

on september 5, xiao xiongwu, the corresponding author of the paper, said in an interview with jimu news that after more than five months of research and continuous satellite remote sensing tracking, the research team used powerful artificial intelligence and satellite remote sensing technology to record the tragic state of the gaza strip being attacked by missiles step by step with detailed data and precise charts. xiao xiongwu said that the research results provide an important reference for the intelligent monitoring and accurate assessment of various natural disasters and conflicts in the global context.

academician li deren guides the team to carry out technical research (source: interviewee)

public information shows that in this study, the research team combined artificial intelligence with the high-resolution satellite remote sensing of luojia-3-01 independently developed by the laboratory, and proposed a dynamic monitoring method for remote sensing of war damage to multi-category urban objects based on high-precision time series change detection. it achieved near-real-time detection of missile craters and damaged buildings in the gaza strip, as well as quantitative assessment of war damage to different categories of objects in different provinces and on different dates, and automatic assessment of damage levels of different categories of buildings.

on the evening of the 5th, xiao xiongwu told jimu news that the research project will be conducted from october 17, 2023 to march 2, 2024. during this period, every four days or so, wuhan university's luojia-3-01 satellite will come over the gaza strip to take a series of satellite remote sensing photos. after the photos are transmitted back, the research team uses its own artificial intelligence algorithm to process and analyze these photos with a resolution of 0.5 meters.

there are multiple items analyzed, including information such as missile craters, building damage, and farmland damage.

automatic detection results of missile craters (location, size, date and number) and damaged universities in the gaza strip based on time-series satellite remote sensing (source: interviewee)

"the information obtained from our artificial intelligence analysis is quite accurate and comprehensive," said xiao xiongwu. the system's monitoring information on craters is detailed to the size and coordinates of each crater, the exact location of new craters added each month, and the total number of craters. according to statistics from xiao xiongwu's team, a total of 3,747 craters were monitored in the gaza strip during the more than four months of his team's research project. the schematic diagram of the relevant research results shows that the entire gaza strip is covered with dense black spots, all of which are where missiles landed.

there are five provinces in the gaza strip. with this remote sensing dynamic monitoring method, the damage to buildings in each province can be clearly detected every time the satellite takes a photo, and the building type can be identified and determined. "whether the damaged building is a hospital, school, residential building, or a place of worship, it can be accurately determined through artificial intelligence combined with remote sensing photos." xiao xiongwu explained that in addition to the building category, artificial intelligence can also analyze the degree of damage to the building, and classify different buildings into five levels according to the degree of damage: no damage, slight damage, moderate damage, severe damage, and complete damage. this is a problem that artificial intelligence around the world has not yet fully solved, but the program can currently do it.

in the illustrations of the paper, jimu news reporters can clearly see that in october last year, only a few buildings in gaza were damaged, but the degree of damage has continued to rise since then. by march this year, the red dots of damaged buildings were almost everywhere in the territory. the damaged buildings accounted for 58.4% of the entire territory.

analysis of building damage and trends in five provinces of the gaza strip at different times (source: interviewee)

xiao xiongwu said that the results of the project directly show the density of missile bombings in the gaza strip, the damage to different types of buildings, the degree of damage and the damage to farmland. it can let the outside world know the real situation of civilians in the gaza strip, such as the poor living environment and food shortage. the research team also emphasized the need for an immediate ceasefire through these first survey results, and strongly called on the international community to provide humanitarian assistance and post-war reconstruction to the gaza strip.

wuhan university's luojia-3-01 satellite was launched into space in january last year, mainly for scientific experiments. its main function is to obtain high-resolution data and observe the world. the characteristics of this research result, in addition to taking photos with the help of satellite remote sensing, focus on the comprehensiveness, accuracy and automation of image information extraction and analysis brought about by the powerful artificial intelligence algorithm of its team.

xiao xiongwu believes that the results of this study provide an important reference for the intelligent monitoring and accurate assessment of various natural disasters, not only in conflict areas.

(source: jimu news)

for more exciting information, please download the "jimu news" client from the application market. please do not reprint without authorization. you are welcome to provide news clues, and you will be paid once adopted.