Performance prediction for generation of point clouds from passive imagery

A system of generating point clouds from passive images, comprising:

• a non-transitory computer-readable storage medium storing instructions for execution by one or more processors;
• and
• one or more processors that, when the instructions are executed, are configured to:
• receive a first plurality of passive images, each of the passive images from the first plurality of passive images corresponding to a geographic location;
• select a reference image from the first plurality of passive images;
• chip the first plurality of passive images using the reference image to obtain a second plurality of passive images, the second plurality of passive images being a subset of the first plurality of passive images;
• form a first plurality of image clusters, wherein each image cluster includes two or more passive images selected from the second plurality of passive images;
• perform a metadata-based performance prediction on each image cluster of the first plurality of image clusters to obtain a corresponding metadata-based performance quality score for each image cluster, the metadata-based performance quality score indicating the predicted quality of a point cloud generated from the corresponding image cluster;
• select a second plurality of image clusters from the first plurality of image clusters based on the metadata-based performance quality score;
• perform a correlator-based performance prediction on each image cluster of the second plurality of image clusters to obtain a corresponding correlator-based performance quality score for each image cluster of the second plurality of image clusters, the correlator-based performance quality score indicating the predicted quality of a point cloud generated from the corresponding image cluster;
• select a third plurality of image clusters from the second plurality of image clusters based on the correlator-based performance quality score;
• generate a point cloud for each of the image clusters of the third plurality of image clusters to obtain a plurality of point clouds, each point cloud comprising a plurality of points, each point being defined by at least three coordinates of a three-dimensional coordinate system;
• select one or more point clouds from the plurality of point clouds based on at least one quantitative quality metric;
• and
• output the selected one or more point clouds to a designated location in a selected electronic format.

A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors, cause the one or more processors to perform a plurality of operations, the operations comprising:

• receiving a first plurality of passive images, each of the passive images from the first plurality of passive images corresponding to a geographic location;
• selecting a reference image from the first plurality of passive images;
• chipping the first plurality of passive images using the reference image to obtain a second plurality of passive images, the second plurality of passive images being a subset of the first plurality of passive images;
• forming a first plurality of image clusters, wherein each image cluster includes two or more passive images selected from the second plurality of passive images;
• performing a metadata-based performance prediction on each image cluster of the first plurality of image clusters to obtain a corresponding metadata-based performance quality score for each image cluster, the metadata-based performance quality score indicating the predicted quality of a point cloud generated from the corresponding image cluster;
• selecting a second plurality of image clusters from the first plurality of image clusters based on the metadata-based performance quality score;
• performing a correlator-based performance prediction on each image cluster of the second plurality of image clusters to obtain a corresponding correlator-based performance quality score for each image cluster of the second plurality of image clusters, the correlator-based performance quality score indicating the predicted quality of a point cloud generated from the corresponding image cluster;
• selecting a third plurality of image clusters from the second plurality of image clusters based on the correlator-based performance quality score;
• generating a point cloud for each of the image clusters of the third plurality of image clusters to obtain a plurality of point clouds, each point cloud comprising a plurality of points, each point being defined by at least three coordinates of a three-dimensional coordinate system;
• selecting one or more point clouds from the plurality of point clouds based on at least one quantitative quality metric;
• and
• outputting the selected one or more point clouds to a designated location in a selected electronic format.