Mapping and tracking system with features in three-dimensional space
First Claim
1. A method for real-time tracking of features, the method comprising:
- acquiring a collection of images and one or more subsequent collections of images to create a three-dimensional space, wherein the collection of images is composed of a first plurality of images captured substantially simultaneously from multiple sources and the one or more subsequent collections of images are composed of a second plurality of images captured substantially simultaneously from the multiple sources, and wherein the second plurality of images are captured later in time than the first plurality of images;
extracting features in each image among the first plurality of images, wherein each feature is a region of multiple pixels within the corresponding image, each region having a feature descriptor vector describing the region;
applying a refinement routine to each feature descriptor vector to identify sub-features indicative of a prominent point within each feature, wherein when a sub-feature cannot be identified within a feature descriptor vector, the corresponding feature is deleted;
identifying a candidate feature in the three-dimensional space for tracking; and
tracking the candidate feature in the three-dimensional space over time using the sub-features within the second plurality of images, wherein the tracked candidate feature is used to determine a travelled path of the multiple sources.
2 Assignments
0 Petitions
Accused Products
Abstract
LK-SURF, Robust Kalman Filter, HAR-SLAM, and Landmark Promotion SLAM methods are disclosed. LK-SURF is an image processing technique that combines Lucas-Kanade feature tracking with Speeded-Up Robust Features to perform spatial and temporal tracking using stereo images to produce 3D features can be tracked and identified. The Robust Kalman Filter is an extension of the Kalman Filter algorithm that improves the ability to remove erroneous observations using Principal Component Analysis and the X84 outlier rejection rule. Hierarchical Active Ripple SLAM is a new SLAM architecture that breaks the traditional state space of SLAM into a chain of smaller state spaces, allowing multiple tracked objects, multiple sensors, and multiple updates to occur in linear time with linear storage with respect to the number of tracked objects, landmarks, and estimated object locations. In Landmark Promotion SLAM, only reliable mapped landmarks are promoted through various layers of SLAM to generate larger maps.
44 Citations
17 Claims
-
1. A method for real-time tracking of features, the method comprising:
-
acquiring a collection of images and one or more subsequent collections of images to create a three-dimensional space, wherein the collection of images is composed of a first plurality of images captured substantially simultaneously from multiple sources and the one or more subsequent collections of images are composed of a second plurality of images captured substantially simultaneously from the multiple sources, and wherein the second plurality of images are captured later in time than the first plurality of images; extracting features in each image among the first plurality of images, wherein each feature is a region of multiple pixels within the corresponding image, each region having a feature descriptor vector describing the region; applying a refinement routine to each feature descriptor vector to identify sub-features indicative of a prominent point within each feature, wherein when a sub-feature cannot be identified within a feature descriptor vector, the corresponding feature is deleted; identifying a candidate feature in the three-dimensional space for tracking; and tracking the candidate feature in the three-dimensional space over time using the sub-features within the second plurality of images, wherein the tracked candidate feature is used to determine a travelled path of the multiple sources. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
Specification