Image pattern matching utilizing discrete curve matching with a mapping operator
First Claim
1. A method for determining the presence of an object of interest in a target image, the method comprising:
- mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator;
acquiring the target image;
determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image, and wherein said determining the target image discrete curve from the target image comprises;
determining an initial target image discrete curve from the target image; and
mapping the initial target image discrete curve to a mapped target image discrete curve using the mapping operator, wherein the mapped target image discrete curve comprises the target image discrete curve;
performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and
generating pattern matching results based on the distances.
5 Assignments
0 Petitions
Accused Products
Abstract
System and method for determining the presence of an object of interest in a target image. Regions of a target image may be located that match an object of interest, e.g., in a template image, with respect to various information, e.g., luminance, color and/or other types of boundary information. The invention includes improved methods for mapping point sequences (e.g., pixel sequences) or curves to new point sets or curves for curve matching. The method determines the presence of an object of interest in a target image despite of or using various types of topological transformations of the object of interest in the target image. One or more mapping operators are determined based on template curves and/or example target curves. Pattern matching is performed on one or more target images using the mapping operator(s) to generate pattern matching results, and the pattern matching results output.
30 Citations
106 Claims
-
1. A method for determining the presence of an object of interest in a target image, the method comprising:
-
mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image, and wherein said determining the target image discrete curve from the target image comprises; determining an initial target image discrete curve from the target image; and mapping the initial target image discrete curve to a mapped target image discrete curve using the mapping operator, wherein the mapped target image discrete curve comprises the target image discrete curve; performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and generating pattern matching results based on the distances. - 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, 34, 35, 36, 37, 38)
-
-
39. A method for determining the presence of an object of interest in a target image, the method comprising:
-
mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator, wherein the mapped template image discrete curve corresponds to an object of interest; acquiring the target image; determining a plurality of target image discrete curves from the target image, wherein each of the target image discrete curves corresponds to a respective object in the target image; for each of the target image discrete curves, mapping the target image discrete curve to a mapped target image discrete curve using a mapping operator; performing shape matching on each of the mapped target image discrete curves and the mapped template image discrete curve corresponding to the object of interest, wherein said performing shape matching computes a distance for the each of the mapped target image discrete curves relative to the mapped template image discrete curve; and generating pattern matching results based on the distances. - View Dependent Claims (40)
-
-
41. A method for performing curve matching, the method comprising:
-
mapping one or more template image discrete curves to one or more mapped template image discrete curves using respective operators, wherein the one or more mapped template image discrete curves are different from the one or more template image discrete curves, and wherein the one or more mapped template image discrete curves are generated based on an object of interest; acquiring a target image; and determining one or more target image discrete curves from the target image, wherein the one or more target image discrete curves represent an image object in the target image, wherein said determining the one or more target image discrete curves from the target image comprises; determining one or more initial target image discrete curves from the target image; and mapping the one or more initial target image discrete curves to a respective one or more mapped target image discrete curves using the respective mapping operators, wherein the one or more mapped target image discrete curves comprises the one or more target image discrete curves; performing curve matching between the one or more target image discrete curves and the one or more mapped template image discrete curves, wherein said performing curve matching computes a distance measure for the one or more target image discrete curves relative to the one or more mapped template image discrete curves, and wherein said performing curve matching between the one or more target image discrete curves and the one or more mapped template image discrete curve comprises performing curve matching between each of the one or more mapped target image discrete curves and the one or more mapped template image discrete curves; and generating curve matching results based on the distance measures.
-
-
42. A method for determining the presence of an object of interest in a target image, the method comprising:
-
mapping a first shape to a different second shape using an operator, wherein the second shape is different from the first shape, and wherein the second shape is generated based on the object of interest; acquiring the target image; determining one or more third shapes from the target image, wherein each of the one or more third shapes represents a respective image object in the target image, wherein said determining the one more third shapes from the target image comprises; determining one or more initial target shapes from the target image; and mapping the one or more initial target shapes to a respective one or more mapped target shapes using the operator, wherein the one or more mapped target shapes comprises the one or more third shapes; performing curve matching between each of the third shapes and the second shape, wherein said performing curve matching computes a distance measure for each of the third shapes relative to the second shape, wherein said performing curve matching between each of the third shapes and the second shape comprises performing curve matching on each of the one or more mapped target shapes and the second shape; and generating curve matching results based on the distance measures.
-
-
43. A computer readable memory medium which stores program instructions for determining the presence of an object of interest in a target image, wherein the program instructions are executable to perform:
-
mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image, and wherein said determining the target image discrete curve from the target image comprises; determining an initial target image discrete curve from the target image; and mapping the initial target image discrete curve to a mapped target image discrete curve using the mapping operator wherein the mapped target image discrete curves comprises the target image discrete curve; performing curve matching on the target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing curve matching computes a distance for the target image discrete curve relative to the mapped template image discrete curve, wherein said performing curve matching between the target image discrete curve and the mapped template image discrete curve comprises performing curve matching between the mapped target image discrete curve and the mapped template image discrete curve; and generating pattern matching results based on the distances. - View Dependent Claims (44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76)
-
-
77. A computer readable memory medium which stores program instructions for performing curve matching between target image discrete curves and a template image discrete curve, wherein the program instructions are executable to perform:
-
mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator, wherein the mapped template image discrete curve corresponds to an object of interest; acquiring a target image; and determining a plurality of target image discrete curves from the target image, wherein each of the target image discrete curves corresponds to a respective object in the target image, and wherein said determining the plurality of target image discrete curves from the target image comprises; determining a plurality of initial target image discrete curves from the target image; and mapping the plurality of initial target image discrete curves to a respective plurality of mapped target image discrete curves using the mapping operator, wherein the plurality of mapped target image discrete curves comprises the plurality of target image discrete curves; for each of the plurality of target image discrete curves, performing curve matching on the target image discrete curve and the mapped template image discrete curve, wherein said performing curve matching computes a distance for the each of the target image discrete curves relative to the mapped template image discrete curve, and wherein said performing curve matching on each of the target image discrete curves and the mapped template image discrete curve comprises performing curve matching on each of the mapped target image discrete curve and the mapped template image discrete curve; and generating curve matching results based on the distances.
-
-
78. A system for determining the presence of an object of interest in a target image, the system comprising:
-
a computer system, comprising; a processor; and a memory medium coupled to the processor; and an image acquisition device coupled to the processor and the memory medium; wherein the image acquisition device is operable to; acquire the target image; and wherein the memory medium stores program instructions which are executable by the processor to; map a template image discrete curve to a mapped template image discrete curve using a mapping operator; determine a plurality of target image discrete curves from the target image, wherein each of the target image discrete curves corresponds to a respective object in the target image, and wherein, to determine the plurality of target image discrete curves from the target image, the program instructions are further executable by the processor to; determine a plurality of initial target image discrete curves from the target image; and map the plurality of initial target image discrete curves to a respective plurality of mapped target image discrete curves using the mapping operator, wherein the plurality of mapped target image discrete curves comprises the plurality of target image discrete curves; for each of the target image discrete curves, perform curve matching on each of the mapped target image discrete curves and the mapped template image discrete curve corresponding to the object of interest, wherein said performing curve matching computes a distance for the each of the mapped target image discrete curves relative to the mapped template image discrete curve, and wherein, to perform curve matching on each of the target image discrete curves and the mapped template image discrete curve the program instructions are further executable by the processor to; perform curve matching on each of the mapped target image discrete curves and the mapped template image discrete curve; and generate curve matching results based on the distances.
-
-
79. A system for determining the presence of an object of interest in a target image, the system comprising:
-
a computer system, comprising; a processor; and a memory medium coupled to the processor; and an image acquisition device coupled to the processor and the memory medium; wherein the image acquisition device is operable to; acquire the target image; and wherein the memory medium stores program instructions which are executable by the processor to; map one or more template image discrete curves to one or more mapped template image discrete curves using respective operators, wherein the one or more mapped template image discrete curves are different from the one or more template image discrete curves; determine one or more target image discrete curves from the target image, wherein the one or more target image discrete curves represents an image object in the target image, and wherein, to determine the one or more target image discrete curves from the target image, the program instructions are further executable by the processor to; determine one or more initial target image discrete curves from the target image; and map the one or more initial target image discrete curves to a respective one or more mapped target image discrete curves using the respective operators, wherein the one or more mapped target image discrete curves comprises the one or more target image discrete curves; perform curve matching between the one or more target image discrete curves and the one or more mapped template image discrete curves, wherein the one or more mapped template image discrete curves are generated based on the object of interest, wherein said performing curve matching computes a distance measure for the one or more target image discrete curves relative to the one or more mapped template image discrete curves, and wherein, to perform curve matching between the one or more target image discrete curves and the one or more mapped template image discrete curves, the program instructions are further executable by the processor to; perform curve matching between each of the one or more mapped target image discrete curves and the one or more mapped template image discrete curves; and generate curve matching results based on the distance measures.
-
-
80. A system for determining the presence of an object of interest in a target image, the system comprising:
-
a computer system, comprising; a processor; and a memory medium coupled to the processor; and an image acquisition device coupled to the processor and the memory medium; wherein the image acquisition device is operable to; acquire the target image; and wherein the memory medium stores program instructions which are executable by the processor to; map a first shape to different second shape using an operator, wherein the second shape is different from the first shape; determine one or more third shapes from the target image, wherein each of the one or more third shapes represents a respective image object in the target image, and wherein, in determining the one or more third shapes from the target image, the program instructions are further executable by the processor to; determine one or more initial target shapes from the target image; and map the one or more initial target shapes to a respective one or more mapped target shapes using the operator, wherein the one or more mapped target shapes comprise the one or more third shapes; perform curve matching between the second shape and each of the third shapes, wherein the second shape is generated based on the object of interest, wherein said performing curve matching computes a distance measure for each of the third shapes relative to the second shape, and wherein, to perform curve matching between the second shape and each of the third shapes, the program instructions are further executable by the processor to; perform curve matching on each of the one or more mapped target shapes and the second shape; and generate curve matching results based on the distance measures.
-
-
81. A system for determining the presence of an object of interest in a target image, the system comprising:
-
means for mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator; means for acquiring the target image; means for determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image, and wherein said means for determining the target image discrete curve from the target image comprise; means for determining an initial target image discrete curve from the target image; and means for mapping the initial target image discrete curve to a mapped target image discrete curve using the mapping operator, wherein the mapped target image discrete curve comprises the target image discrete curve; means for performing curve matching on the target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing curve matching computes a distance for the target image discrete curve relative to the mapped template image discrete curve, and wherein said means for performing curve matching on the target image discrete curve and the mapped template image discrete curve comprises means for performing curve matching on the mapped target image discrete curve and the mapped template image discrete curve; and means for generating curve matching results based on the distances.
-
-
82. A system for determining the presence of an object of interest in a target image, the system comprising:
-
means for mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator; means for acquiring the target image; means for determining a plurality of target image discrete curves from the target image, wherein each of the target image discrete curves corresponds to a respective object in the target image, and wherein said means for determining the plurality of target image discrete curves from the target image comprises; means for determining a plurality of initial target image discrete curves from the target image; and means for mapping the plurality of initial target image discrete curves to a respective plurality of mapped target image discrete curves using the mapping operator, wherein the plurality of mapped target image discrete curves comprises the plurality of target image discrete curves; means for performing curve matching on each of the target image discrete curves and the mapped template image discrete curve corresponding to the object of interest, wherein said performing curve matching computes a distance for the each of the target image discrete curves relative to the mapped template image discrete curve, and wherein said means for performing curve matching on each of the target image discrete curves and the mapped template image discrete curve comprises means for performing curve matching on each of the mapped target image discrete curves and the mapped template image discrete curve; and means for generating curve matching results based on the distances.
-
-
83. A system for determining the presence of an object of interest in a target image, the system comprising:
-
means for mapping one or more template image discrete curves to one or more mapped template image discrete curves using a respective one or more operators, wherein the one or more mapped template image discrete curves are different from the one or more template image discrete curves; means for acquiring the target image; means for determining one or more target image discrete curves from the target image, wherein the one or more target image discrete curves represent an image object in the target image, and wherein, said means for determining the one or more target image discrete curves from the target image comprises; means for determining one or more initial target image discrete curves from the target image; and means for mapping the one or more initial target image discrete curves to respective one or more mapped target image discrete curves using the respective one or more operators, wherein the one or more mapped target image discrete curves comprise the one or more target image discrete curves; means for performing curve matching between the one or more target image discrete curves and the one or more mapped template image discrete curves, wherein the one or more mapped template image discrete curves are generated based on the object of interest, wherein said performing curve matching computes a distance measure for the one or more target image discrete curves relative to the one or more mapped template image discrete curves, and wherein said means for performing curve matching between the one or more target image discrete curves and the one or more mapped template image discrete curves comprises; means for performing curve matching between each of the one or more mapped target image discrete curves and the one or more mapped template image discrete curves; and means for generating curve matching results based on the distance measures.
-
-
84. A method for characterizing an object of interest for a pattern matching application, the method comprising:
-
receiving data containing the object of interest; determining a template image discrete curve from the data, wherein the template image discrete curve corresponds to the object of interest in the data; and analyzing the object of interest and one or more example target images to determine one or more terms for a mapping operator, wherein the mapping operator comprises a weight vector; and mapping the template image discrete curve to a mapped template image discrete curve using a mapping operator, wherein said mapping comprises; applying the weight vector terms respectively to corresponding points in the template image discrete curve; wherein the mapped template image discrete curve is useable to perform pattern matching to locate instances of the object of interest in acquired target data. - View Dependent Claims (85, 86, 87, 88, 89, 90, 91, 92)
-
-
93. A method for characterizing an object of interest for a pattern matching application, the method comprising:
-
receiving first information regarding the object of interest; determining one or more template image discrete curves from the first information, wherein the one or more template image discrete curves correspond to the object of interest in a template image; receiving second information regarding objects not of interest; determining one or more background object discrete curves from the second information, wherein the one or more background object discrete curves correspond to one or more objects not of interest; computing the mapping operator from the one or more template image discrete curves and the one or more background object discrete curves; wherein the mapping operator operates to amplify differences between each of the discrete curves; and mapping the one or more template image discrete curves to a corresponding one or more mapped template image discrete curves using the mapping operator; wherein the one or more mapped template image discrete curves are useable to perform pattern matching to locate instances of the object of interest in acquired target data. - View Dependent Claims (94, 95, 96)
-
-
97. A method for determining the presence of an object of interest in a target image, the method comprising:
-
mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image, wherein said determining a target image discrete curve from the target image comprises; detecting at least one boundary present in the target image, wherein the at least one boundary corresponds to a shape of an object in the target image; and determining the target image discrete curve from the at least one boundary; normalizing the target image discrete curve, wherein said normalizing comprises normalizing one or more of; energy of the target image discrete curve; and length of the target image discrete curve; mapping the target image discrete curve to a mapped target image discrete curve using the mapping operator; performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and generating pattern matching results based on the distances.
-
-
98. A method for determining the presence of an object of interest in a target image, the method comprising:
-
mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image, wherein said determining a target image discrete curve from the target image comprises; detecting at least one boundary present in the target image, wherein the at least one boundary corresponds to a shape of an object in the target image; and determining the target image discrete curve from the at least one boundary; mapping the target image discrete curve to a mapped target image discrete curve using the mapping operator; performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and generating pattern matching results based on the distances.
-
-
99. A method for determining the presence of an object of interest in a target image, the method comprising:
-
analyzing one or more example target images to determine one or more example target image discrete curves, wherein each of the one or more example target image discrete curves corresponds to an image object or part of an image object in the one or more example target images; characterizing an object of interest to produce a mapped template image discrete curve, wherein said characterizing the object of interest comprises; determining the template image discrete curve from information regarding the object of interest, wherein the template image discrete curve corresponds to the object of interest; determining a mapping operator based on a determined template image discrete curve and the one or more example target image discrete curves mapping the template image discrete curve to the mapped template image discrete curve using the mapping operator; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image; mapping the target image discrete curve to a mapped target image discrete curve using the mapping operator; performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and generating pattern matching results based on the distances.
-
-
100. A method for determining the presence of an object of interest in a target image, the method comprising:
-
characterizing an object of interest to produce a mapped template image discrete curve, wherein said characterizing the object of interest comprises; determining a template image discrete curve from information regarding the object of interest, wherein the template image discrete curve corresponds to the object of interest; mapping a template image discrete curve to the mapped template image discrete curve using a mapping operator; analyzing the mapped template image discrete curve and one or more example target image discrete curves generated from one or more example target images to determine at least one parameter for the mapping operator; and configuring the mapping operator according to the determined at least one parameter; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image; mapping the target image discrete curve to a mapped target image discrete curve using the mapping operator; performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and generating pattern matching results based on the distances.
-
-
101. A method for determining the presence of an object of interest in a target image, the method comprising:
-
receiving first information regarding an object of interest; receiving second information regarding one or more objects that are not of interest; determining at least one parameter for a mapping operator based on the first information and the second information, wherein the at least one parameter is determined to enhance a distance between mapped discrete curves corresponding to an object of interest and mapped discrete curves corresponding to the one or more objects that are not of interest; and configuring the mapping operator according to the determined at least one parameter mapping a template image discrete curve to a mapped template image discrete curve using the mapping operator; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image; mapping the target image discrete curve to a mapped target image discrete curve using the mapping operator; performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and generating pattern matching results based on the distances.
-
-
102. A method for determining the presence of an object of interest in a target image, the method comprising:
-
mapping a template image discrete curve to a mapped template image discrete curve using a mapping operator, wherein the mapping operator comprises a weight vector w, wherein the template image discrete curve comprises a first vector, wherein said weight vector w, and said first vector have the same number of elements, and wherein said mapping the template image discrete curve to the mapped template image discrete curve comprises; multiplying each element of the weight vector w with corresponding elements in the first vector to generate a second vector, wherein elements of the second vector comprise the mapped template image discrete curve; acquiring the target image; determining a target image discrete curve from the target image, wherein the target image discrete curve corresponds to a respective object in the target image; mapping the target image discrete curve to a mapped target image discrete curve using the mapping operator; performing shape matching on the mapped target image discrete curve and the mapped template image discrete curve, wherein the mapped template image discrete curve corresponds to the object of interest, wherein said performing shape matching computes a distance for the mapped target image discrete curve relative to the mapped template image discrete curve; and generating pattern matching results based on the distances.
-
-
103. A method for characterizing an object of interest for a pattern matching application, the method comprising:
-
receiving data containing the object of interest; determining a template image discrete curve from the data, wherein the template image discrete curve corresponds to the object of interest in the data; and analyzing the object of interest and one or more example target images to determine at least one parameter for the mapping operator; configuring the mapping operator according to the determined at least one parameter; and mapping the template image discrete curve to a mapped template image discrete curve using the configured mapping operator; wherein the mapped template image discrete curve is useable to perform pattern matching to locate instances of the object of interest in acquired target data. - View Dependent Claims (104)
-
-
105. A method for characterizing an object of interest for a pattern matching application, the method comprising:
-
receiving data containing the object of interest; receiving first information regarding the object of interest; receiving second information regarding one or more objects that are not of interest; determining at least one parameter for a mapping operator based on the first information and the second information; and configuring the mapping operator according to the determined at least one parameter determining a template image discrete curve from the data, wherein the template image discrete curve corresponds to the object of interest in the data; and mapping the template image discrete curve to a mapped template image discrete curve using the mapping operator; wherein the mapped template image discrete curve is useable to perform pattern matching to locate instances of the object of interest in acquired target data. - View Dependent Claims (106)
-
Specification