Surface fractures have a great impact on ice shelf stability in Antarctica and can be considered precursors of ice shelf calving.However,our understanding of the spatial distribution and temporal evolution of surface ...Surface fractures have a great impact on ice shelf stability in Antarctica and can be considered precursors of ice shelf calving.However,our understanding of the spatial distribution and temporal evolution of surface fractures on the Antarctic ice shelf is limited.In this study,a ResUNet model was implemented on the Moderate Resolution Imaging Spectroradiometer(MODIS)-based Mosaic of Antarctica(MOA)to identify the spatial distribution of Antarctic ice shelf surface fractures in 2004,2009,and 2014.The accuracy of identification had an F1 value of 0.771.Our model identified 44744.59±2619.61 km^(2)of surface fractures in 2004,43737.15±2644.60 km^(2)in 2009,and 42978.67±2639.33 km^(2)in 2014.The reduction is primarily attributed to the variation in surface fractures within 20 km of the ice front,paratactically in the Amundsen and Wilkes sectors.Ice shelves in the Amundsen sector typically have a widespread distribution of surface fractures,with particularly high concentrations found in the Thwaites Ice Shelf,Crosson Ice Shelf and Getz Ice Shelf.The Brunt Ice Shelf also exhibits numerous surface fractures.This study provides comprehensive and detailed information about surface fractures on Antarctic ice shelves,carrying implications for evaluating ice shelf vulnerability.展开更多
Landslide detection is a hot topic in the remote sensing community,particularly with the current rapid growth in volume(and variety)of Earth observation data and the substantial progress of computer vision.Deep learni...Landslide detection is a hot topic in the remote sensing community,particularly with the current rapid growth in volume(and variety)of Earth observation data and the substantial progress of computer vision.Deep learning algorithms,especially fully convolutional networks(FCNs),and variations like the ResU-Net have been used recently as rapid and automatic landslide detection approaches.Although FCNs have shown cutting-edge results in automatic landslide detection,accuracy can be improved by adding prior knowledge through possible frameworks.This study evaluates a rulebased object-based image analysis(OBIA)approach built on probabilities resulting from the ResU-Net model for landslide detection.We train the ResU-Net model using a landslide dataset comprising landslide inventories from various geographic regions,including our study area and test the testing area not used for training.In the OBIA stage,we frst calculate land cover and image difference indices for pre-and post-landslide multi-temporal images.Next,we use the generated indices and the resulting ResU-Net probabilities for image segmentation;the extracted landslide object candidates are then optimized using rule-based classification.In the result validation section,the landslide detection of the proposed integration of the ResU-Net with a rule-based classification of OBIA is compared with that of the ResU-Net alone.Our proposed approach improves the mean intersection-over-union of the resulting map from the ResU-Net by more than 22%.展开更多
Liver tumor is the fifth most occurring type of tumor in men and the ninth most occurring type of tumor in women according to recent reports of Global cancer statistics 2018.There are several imaging tests like Comput...Liver tumor is the fifth most occurring type of tumor in men and the ninth most occurring type of tumor in women according to recent reports of Global cancer statistics 2018.There are several imaging tests like Computed Tomography(CT),Magnetic Resonance Imaging(MRI),and ultrasound that can diagnose the liver tumor after taking the sample from the tissue of the liver.These tests are costly and time-consuming.This paper proposed that image processing through deep learning Convolutional Neural Network(CNNs)ResUNet model that can be helpful for the early diagnose of tumor instead of conventional methods.The existing studies have mainly used the two Cascaded CNNs for liver segmentation and evaluation of Region Of Interest(ROI).This study uses ResUNet,an updated version of U-Net and ResNet Models that utilize the service of Residential blocks.We apply over method on the 3D-IRCADb01 dataset that is based on CT slices of liver tumor affected patients.The results showed the True Value Accuracy around 99%and F1 score performance around 95%.This method will be helpful for early and accurate diagnose of the Liver tumor to save the lives of many patients in the field of Biotechnology.展开更多
基金supported by the National Natural Science Foundation of China(grant no.41830536,41925027)the Innovation Group Project of Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai)(grant no.311022003)the Fundamental Research Funds for the Central Universities,Sun Yat-sen University(23ptpy99,231GBJ022).
文摘Surface fractures have a great impact on ice shelf stability in Antarctica and can be considered precursors of ice shelf calving.However,our understanding of the spatial distribution and temporal evolution of surface fractures on the Antarctic ice shelf is limited.In this study,a ResUNet model was implemented on the Moderate Resolution Imaging Spectroradiometer(MODIS)-based Mosaic of Antarctica(MOA)to identify the spatial distribution of Antarctic ice shelf surface fractures in 2004,2009,and 2014.The accuracy of identification had an F1 value of 0.771.Our model identified 44744.59±2619.61 km^(2)of surface fractures in 2004,43737.15±2644.60 km^(2)in 2009,and 42978.67±2639.33 km^(2)in 2014.The reduction is primarily attributed to the variation in surface fractures within 20 km of the ice front,paratactically in the Amundsen and Wilkes sectors.Ice shelves in the Amundsen sector typically have a widespread distribution of surface fractures,with particularly high concentrations found in the Thwaites Ice Shelf,Crosson Ice Shelf and Getz Ice Shelf.The Brunt Ice Shelf also exhibits numerous surface fractures.This study provides comprehensive and detailed information about surface fractures on Antarctic ice shelves,carrying implications for evaluating ice shelf vulnerability.
基金funded by the Institute of Advanced Research in Artificial Intelligence(IARAl)GmbHInstitute of Advanced Research in Artificial Intelligence(IARAl)GmbH Address:LandstraBer HauptstraBe 5,1030 Vienna,Austria[VAT number(UID):ATU74131236].
文摘Landslide detection is a hot topic in the remote sensing community,particularly with the current rapid growth in volume(and variety)of Earth observation data and the substantial progress of computer vision.Deep learning algorithms,especially fully convolutional networks(FCNs),and variations like the ResU-Net have been used recently as rapid and automatic landslide detection approaches.Although FCNs have shown cutting-edge results in automatic landslide detection,accuracy can be improved by adding prior knowledge through possible frameworks.This study evaluates a rulebased object-based image analysis(OBIA)approach built on probabilities resulting from the ResU-Net model for landslide detection.We train the ResU-Net model using a landslide dataset comprising landslide inventories from various geographic regions,including our study area and test the testing area not used for training.In the OBIA stage,we frst calculate land cover and image difference indices for pre-and post-landslide multi-temporal images.Next,we use the generated indices and the resulting ResU-Net probabilities for image segmentation;the extracted landslide object candidates are then optimized using rule-based classification.In the result validation section,the landslide detection of the proposed integration of the ResU-Net with a rule-based classification of OBIA is compared with that of the ResU-Net alone.Our proposed approach improves the mean intersection-over-union of the resulting map from the ResU-Net by more than 22%.
基金The authors extend their appreciation to the Deanship of Scientific Research at King Saud University for funding this work through research group No RG-1438-089.
文摘Liver tumor is the fifth most occurring type of tumor in men and the ninth most occurring type of tumor in women according to recent reports of Global cancer statistics 2018.There are several imaging tests like Computed Tomography(CT),Magnetic Resonance Imaging(MRI),and ultrasound that can diagnose the liver tumor after taking the sample from the tissue of the liver.These tests are costly and time-consuming.This paper proposed that image processing through deep learning Convolutional Neural Network(CNNs)ResUNet model that can be helpful for the early diagnose of tumor instead of conventional methods.The existing studies have mainly used the two Cascaded CNNs for liver segmentation and evaluation of Region Of Interest(ROI).This study uses ResUNet,an updated version of U-Net and ResNet Models that utilize the service of Residential blocks.We apply over method on the 3D-IRCADb01 dataset that is based on CT slices of liver tumor affected patients.The results showed the True Value Accuracy around 99%and F1 score performance around 95%.This method will be helpful for early and accurate diagnose of the Liver tumor to save the lives of many patients in the field of Biotechnology.
