12/8/2023 0 Comments Adversarial network radarThe issue of detecting faint layers was improved in based on an electric charged particle concept. However, their technique has difficulty in detecting the ice bottom when it is faint. Therefore, it does not need any manual re-initialization and was automatically applied to a large dataset. This problem was fixed in, where authors introduced a distance regularization term in the level set approach to maintain the regularity of level set intrinsically. However, for every single image, the user needs to re-initialize the curve manually and as a result, the method is quite slow and was applied only to a small dataset. used a level set technique for estimating bedrock and surface layers. The extension of this work was presented in where they used Markov-chain Monte Carlo to sample from the joint distribution over all possible layers conditioned on an image. used probabilistic graphical models for detecting the ice layer boundary in echogram images from Greenland and Antarctica. Several semi-automated and automated methods exist for layer finding and estimating ice thickness in radar images. Radars are one of the most important sensors that can penetrate through ice and give us information about the ice thickness. The shape of the landscape beneath the thick ice sheets is an important factor in predicting ice flow. Precise calculation of ice thickness is very important for sea level and flood monitoring. Melting polar ice sheets and mountain glaciers have considerable influence on sea-level rise (SLR) and ocean currents potential floods in coastal regions could put millions of people around the world at risk. Ice loss in Greenland and Antarctica has accelerated in recent decades. To the best of our knowledge, this is the first work in creating radar sounder imagery based on generative adversarial network. However, the synthetic images generated by GANs cannot be used solely for training a neural network (training on synthetic and testing on real) as they cannot simulate all of the radar characteristics such as noise or Doppler effects. Our experiments show that synthetic radar images generated by generative adversarial network (GAN) can be used in combination with real images for data augmentation and training of deep neural networks. We also tested the quality of a state-of-the-art contour detection algorithm on synthetic data and different combinations of real and synthetic data. We conducted several experiments to test the quality of the generated radar imagery. In this research, we evaluated the performance of synthetically generated radar images based on modified cycle-consistent adversarial networks. In this way, training data can be quickly augmented with additional images. Instead of labeling real images, we can generate synthetic data based on arbitrary labels. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |