Digitalglobe, Incorporated

Georgios Ouzounis

Kostas Stamatiou

Nikki Aldeborgh

A system and methods for unsupervised land use and land cover detection using a classifier that produces a plurality of class image layers which are filtered to remove misclassified same-label pixel groupings, a class resolution module that reduces multiple pixel labels to a single one if applicable and a reconstruction module that generates the output land use and land cover image.

Unsupervised land use and land cover detection

A system and methods for radiometric anomaly detection using non-target clustering, wherein a hierarchy generator organizes the image information content into a hierarchical data representation structure, and a non-target clustering engine processes the hierarchical model to identify large homogeneous regions and significantly dissimilar smaller regions within them based on search criteria.

Anomaly detection using non-target clustering

A system and methods for muddy water detection using normalized semantic layers, wherein a spectrum analyzer isolates spectrum bands within an image to produce a set of three normalized differential index images from which a composite color image is created, from which a power band is computed, from which a two-color image is produced, and then filters image components within the two-color representation based on defined criteria.

Muddy water detection using normalized semantic layers

A system and various methods for processing an image to produce a hierarchical image representation model, segment the image model using shape criteria to produce positive and negative training data sets as well as a search-space data set comprising shapes matched to a search query provided as input, and using the training data sets to train a machine learning model to improve recognition of shapes that are similar to an input query without being exact matches, to improve object recognition.

Shape-based segmentation using hierarchical image representations for automatic training data generation and search space specification for machine learning algorithms

Automatic characterization or categorization of portions of an input multispectral image based on a selected reference multispectral image. Sets (e.g., vectors) of radiometric descriptors of pixels of each component of a hierarchical representation of the input multispectral image can be collectively manipulated to obtain a set of radiometric descriptors for the component. Each component can be labeled as a (e.g., relatively) positive or negative instance of at least one reference multispectral image (e.g., mining materials, crops, etc.) through a comparison of the set of radiometric descriptors of the component and a set of radiometric descriptors for the reference multispectral image. Pixels may be labeled (e.g., via color, pattern, etc.) as positive or negative instances of the land use or type of the reference multispectral image in a resultant image based on components within which the pixels are found.

Multi-spectral image labeling with radiometric attribute vectors of image space representation components

A system for automatically extracting or isolating structures or areas of interest (e.g., built-up structures such as buildings, houses, shelters, tents; agricultural areas; etc.) from HR/VHR overhead imagery data by way of making as little as a single pass through a hierarchical data structure of input image components (where pixels are grouped into components based on any appropriate definition or measure of dissimilarity between adjacent pixels of the input image) to identify candidate components (e.g., possible structures of interest) free of necessarily having to re-iterate the same operator configured with different threshold parameters for a plurality of values.

Growing community of inventors

Longmont, CO, United States of America

Georgios Ouzounis