Anders U. Waldeland
University of Oslo
Signal processingDeep learningGeologyArtificial neural networkConvolutional neural network
What is this?
Publications 9
#1H. ZhaoH-Index: 1
#2Anders U. Waldeland (University of Oslo)H-Index: 3
Last. Einar IversenH-Index: 13
view all 5 authors...
Advanced seismic imaging and inversion are dependent on a velocity model that is sufficiently accurate to render reliable and meaningful results. For that reason, methods for extracting such velocity models from seismic data are always in high demand and are topics of active investigation. Velocity models can be obtained from both the time and depth domains. Relying on the former, time migration is an inexpensive, quick and robust process. In spite of its limitations, especially in the case of c...
#1Olav Brautaset (Norwegian Computing Center)H-Index: 1
#2Anders U. Waldeland (Norwegian Computing Center)H-Index: 3
Last. Nils Olav HandegardH-Index: 20
view all 7 authors...
1 CitationsSource
#1Anders U. Waldeland (University of Oslo)H-Index: 3
#2Tiago Coimbra (University of Oslo)H-Index: 1
Last. Leiv-J. Gelius (State University of Campinas)H-Index: 16
view all 5 authors...
#1Øivind Due Trier (Norwegian Computing Center)H-Index: 6
#2David C. Cowley (Historic Environment Scotland)H-Index: 3
Last. Anders U. Waldeland (Norwegian Computing Center)H-Index: 3
view all 3 authors...
6 CitationsSource
#1Anders U. Waldeland (University of Oslo)H-Index: 3
#2Are Charles Jensen (University of Oslo)H-Index: 9
Last. Anne H. Schistad Solberg (University of Oslo)H-Index: 16
view all 4 authors...
Abstract Deep-learning methods have proved successful recently for solving problems in image analysis and natural language processing. One of these methods, convolutional neural networks (CNNs), is revolutionizing the field of image analysis and pushing the state of the art. CNNs consist of layers of convolutions with trainable filters. The input to the network is the raw image or seismic amplitudes, removing the need for feature/attribute engineering. During the training phase, the filter coeff...
20 CitationsSource
#1Anders U. Waldeland (Norwegian Computing Center)H-Index: 3
#2Jarle Hamar Reksten (Norwegian Computing Center)H-Index: 1
Last. Arnt-Borre Salberg (Norwegian Computing Center)H-Index: 7
view all 3 authors...
Detection of avalanches is critical for keeping avalanche inventories and management of emergency situations. In this paper we propose a deep-learning based avalanche detection method for SAR images. We utilize an existing method for proposing candidate regions, based on change detection in SAR images from multiple passes over the same area. Then a convolutional neural network is used to classify whether the candidate regions contain an avalanche or not. The proposed methodology applies existing...
1 CitationsSource
#1Anders U. Waldeland (University of Oslo)H-Index: 3
#2Hao Zhao (University of Oslo)H-Index: 1
Last. Leiv-J. Gelius (University of Oslo)H-Index: 16
view all 5 authors...
ABSTRACTThe common-reflection-surface (CRS) method offers a stack with higher signal-to-noise ratio at the cost of a time-consuming semblance search to obtain the stacking parameters. We have developed a fast method for extracting the CRS parameters using local slope and curvature. We estimate the slope and curvature with the gradient structure tensor and quadratic structure tensor on stacked data. This is done under the assumption that a stacking velocity is already available. Our method was co...
1 CitationsSource
Traditional methods for salt classification consist of choosing a set of attributes that are sensitive to the characteristics of salt bodies and training a classification algorithm to discriminate between salt and other geological structures. Convolutional neural networks have the advantage of combining attribute extraction and classification in one network. This allows both the attributes and classification to be trainable for the given application. In this work we show how this technique can b...
8 CitationsSource
We present an approach to detect and segment salt bodies in new unlabelled datasets based on 3D attributes and a classifier. The classifier is trained on one labelled dataset and used to classify salt bodies on new unlabelled datasets. Through a forward attribute selection algorithm and manual inspection of attribute images and classified images, we have evaluated a wide set of attributes and classifiers with the aim to be able to detect salt in unlabelled datasets. The simple nearest mean class...