Image-based feature detection using edge vectors
First Claim
1. An automated image-based feature detection apparatus, comprising:
- a digital image capturing device;
a non-transitory computer-readable memory storing feature detection software instructions; and
a processor coupled with the digital image capturing device and the memory, and configured to operate as a feature detector upon execution of the software instructions, wherein the feature detector is configurable to;
obtain, via the digital image capturing device, digital image frames of a real-time video stream comprising digital images of a streetscape including a plurality of objects, wherein the plurality of objects includes at least one vehicle;
detect a plurality of edges associated with at least one object of the plurality of objects within a plurality of frames of the digital image frames;
track, in real-time across the plurality of frames, image features based on one or more anchor points associated with at least one edge of the plurality of edges;
assign a vector for each of at least some points of a plurality of points along the at least one edge such that a plurality of vectors is associated with the at least one edge, wherein each anchor point of the one or more anchor points corresponds to a vector of the plurality of vectors, thereby defining an anchor point vector, and wherein each anchor point is associated with an analysis grid located on the at least one edge;
generate a similarity measure as a function of relative measures between one of the anchor point vectors and at least some of the plurality of vectors within an analysis grid that corresponds to the one of the anchor point vectors; and
enable a networked device to take an action associated with the at least one object as a function of the similarity measure.
3 Assignments
0 Petitions
Accused Products
Abstract
Techniques are provided in which a plurality of edges are detected within a digital image. An anchor point located along an edge of the plurality of edges is selected. An analysis grid associated with the anchor point is generated, the analysis grid including a plurality of cells. An anchor point normal vector comprising a normal vector of the edge at the anchor point is calculated. Edge pixel normal vectors comprising normal vectors of the edge at locations along the edge within the cells of the analysis grid are calculated. A histogram of similarity is generated for each of one or more cells of the analysis grid, each histogram of similarity being based on a similarity measure between each of the edge pixel normal vectors within a cell and the anchor point normal vector, and a descriptor is generated for the analysis grid based on the histograms of similarity.
-
Citations
33 Claims
-
1. An automated image-based feature detection apparatus, comprising:
-
a digital image capturing device; a non-transitory computer-readable memory storing feature detection software instructions; and a processor coupled with the digital image capturing device and the memory, and configured to operate as a feature detector upon execution of the software instructions, wherein the feature detector is configurable to; obtain, via the digital image capturing device, digital image frames of a real-time video stream comprising digital images of a streetscape including a plurality of objects, wherein the plurality of objects includes at least one vehicle; detect a plurality of edges associated with at least one object of the plurality of objects within a plurality of frames of the digital image frames; track, in real-time across the plurality of frames, image features based on one or more anchor points associated with at least one edge of the plurality of edges; assign a vector for each of at least some points of a plurality of points along the at least one edge such that a plurality of vectors is associated with the at least one edge, wherein each anchor point of the one or more anchor points corresponds to a vector of the plurality of vectors, thereby defining an anchor point vector, and wherein each anchor point is associated with an analysis grid located on the at least one edge; generate a similarity measure as a function of relative measures between one of the anchor point vectors and at least some of the plurality of vectors within an analysis grid that corresponds to the one of the anchor point vectors; and enable a networked device to take an action associated with the at least one object as a function of the similarity measure. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32)
-
-
33. An automated system for image-based feature detection, comprising:
-
a digital image capturing device configured to obtain digital image frames of a real-time video stream comprising digital images of a streetscape including a plurality of vehicles; and a feature detector in communication with the digital image capturing device and configured to; detect a plurality of edges associated with at least one vehicle of the plurality of vehicles within a plurality of frames of the digital image frames; track, in real-time across the plurality of frames, image features based on one or more anchor points associated with at least one edge of the plurality of edges; assign a vector for each of at least some points of a plurality of points along the at least one edge such that a plurality of vectors is associated with the at least one edge, wherein each anchor point of the one or more anchor points corresponds to a vector of the plurality of vectors, thereby defining an anchor point vector, and wherein each anchor point is associated with an analysis grid located on the at least one edge; generate a similarity measure as a function of relative measures between one of the anchor point vectors and at least some of the plurality of vectors within an analysis grid that corresponds to the one of the anchor point vectors; and enable a networked device to take an action associated with the at least one vehicle based on the similarity measure.
-
Specification