US 12,307,638 B1
Systems and methods for assessing a presence or absence of celestial objects from a plurality of images of a cosmic space acquired by a plurality of imaging satellites
Frederic Pelletier, Quebec (CA); Daniel O'Connell, Saint Lazarre (CA); Peter Klimas, Ottawa (CA); Jean-Claude Leclerc, Quebec (CA); Narendra Gollu, Montreal (CA); Pierre Tremblay, Quebec (CA); Ryan Comeau, Kelowna (CA); Noemi Giammichele, Montreal (CA); and Priyatharsan Epoor Rajasekar, Luxembourg (LU)
Assigned to NORTHSTAR EARTH & SPACE INC., Montreal (CA)
Filed by NorthStar Earth & Space Inc., Montreal (CA)
Filed on Jan. 13, 2025, as Appl. No. 19/018,267.
Application 19/018,267 is a continuation of application No. 18/740,931, filed on Jun. 12, 2024, granted, now 12,249,055.
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 5/92 (2024.01); G06T 7/80 (2017.01)
CPC G06T 5/92 (2024.01) [G06T 7/80 (2017.01)] 24 Claims
OG exemplary drawing
 
1. A method, comprising:
receiving, by at least one centralized computer processor of at least one centralized computer of a central computing processing arrangement, image data from a plurality of images, acquired by a plurality of imaging satellites;
determining, by the at least one centralized computer processor, a presence or an absence of at least one known object in a series of images from the plurality of images, based at least in part on a temporal behavior of a first plurality of dynamic characteristic image features associated with the at least one known object;
wherein the at least one known object is at least one known RSO, at least one known celestial body, or both; and
determining, by the at least one centralized computer processor, a presence or an absence of at least one unknown object in the series of images, based at least in part on the temporal behavior of a second plurality of dynamic characteristic image features associated with the at least one unknown object by inputting the image data of the series of images into at least one machine learning model that is trained to at least:
group the second plurality of dynamic characteristic image features in the series of images into at least one track, and
determine a likelihood of the at least one track being associated with each of a plurality of registered known objects;
wherein the at least one unknown object is at least one unknown RSO, at least one unknown celestial body, or both;
wherein the plurality of registered known objects comprises a plurality of:
at least one registered known RSOs,
at least one registered known celestial body, or
both; and
modify a data repository based on the presence or the absence of the at least one unknown object in the series of images.