University of Minnesota

Ankush Khandelwal

Vipin Kumar

Anuj Karpatne

Varun Mithal

Shyam Boriah

Xiaowei Jia

Xi C Chen

Karsten Steinhaeuser

Guruprasad Nayak

A method includes receiving a satellite image of an area and classifying each pixel in the satellite image as representing water, land or unknown using a model. For each of a plurality of possible water levels, a cost associated with the water level is determined, wherein determining the cost associated with a water level includes determining a number of pixels for which the model classification must change to be consistent with the water level and determining a difference between the water level and a water level determined for the area at a previous time point. The lowest cost water level is selected and used to reclassify at least one pixel.

Satellite image classification across multiple resolutions and time using ordering constraint among instances

A method includes classifying low-resolution pixels of a low-resolution satellite image of a geographic area to form an initial classification map and selecting at least one physically-consistent classification map of the low-resolution pixels based on the initial classification map. A water level associated with at least one of the physically-consistent classification maps is then used to identify a set of high-resolution pixels representing a perimeter of water in the geographic area.

A method of identifying land cover includes receiving multi-spectral values for a plurality of locations at a plurality of times. A location is selected and for each time in the plurality of times, a latent representation of the multi-spectral values is determined based on a latent representation of multi-spectral values determined for a previous time and multi-spectral values for the previous time of a plurality of other locations that are near the selected location. The determined latent representation is then used to predict a land cover for the selected location at the time.

Predicting land covers from satellite images using temporal and spatial contexts

A method includes receiving data for an entity for each of a plurality of time points. For each of a plurality of time windows that each comprises a respective plurality of time points, a confidence value is determined. The confidence value provides an indication of the degree to which the time window contains data that is useful in discriminating between classes. The confidence values are used to determine a probability of a class and the probability of the class is used to set a predicted class for the entity.

Discovery of shifting patterns in sequence classification

A method for identifying highly-skewed classes using an imperfect annotation of every instance together with a set of features for all instances. The imperfect annotations designate a plurality of instances as belonging to the target rare class and others to the majority class. First, a classifier is trained on the set of features using the imperfect annotation as supervision, to designate each instance to either the rare class or majority class. A combination of the predictions from the trained classifier and the imperfect annotations is then used to classify each instance to either the rare class or majority class. In particular, an instance is classified to the rare class only when both the trained classifier and the imperfect annotation classify the instance to the rare class. Finally, for each instance assigned as a rare class instance by the combination stage, all instances in its neighborhood are re-classified as either rare class or majority class.

Classification of highly-skewed data

A method reduces processing time required to identify locations burned by fire by receiving a feature value for each pixel in an image, each pixel representing a sub-area of a location. Pixels are then grouped based on similarities of the feature values to form candidate burn events. For each candidate burn event, a probability that the candidate burn event is a true burn event is determined based on at least one further feature value for each pixel in the candidate burn event. Candidate burn events that have a probability below a threshold are removed from further consideration as burn events to produce a set of remaining candidate burn events.

Unsupervised spatio-temporal data mining framework for burned area mapping

A method of reducing processing time when assigning geographic areas to land cover labels using satellite sensor values includes a processor receiving a feature value for each pixel in a time series of frames of satellite sensor values, each frame containing multiple pixels and each frame covering a same geographic location. For each sub-area of the geographic location, the sub-area is assigned to one of at least three land cover labels. The processor determines a fraction function for a first sub-area assigned to a first land cover label. The sub-areas that were assigned to the first land cover label are reassigned to one of the second land cover label and the third land cover label based on the fraction functions of the sub-areas.

Unsupervised framework to monitor lake dynamics

A system has an aerial image database containing sensor data representing a plurality of aerial images of an area having multiple sub-areas. A processor applies a classifier to the sensor values to identify a label for each sub-area in each aerial image and to thereby generate an initial label sequence for each sub-area. The processor identifies a most likely land cover state for each sub-area based on the initial label sequence, a confusion matrix and a transition matrix. For each sub-area, the processor stores the most likely land cover state sequence for the sub-area.

Growing community of inventors

Minneapolis, MN, United States of America

Ankush Khandelwal