US2016253574A1PendingUtilityA1

Technologies for determining local differentiating color for image feature detectors

33
Individually held — no corporate assignee on recordPriority: Nov 28, 2013Filed: Nov 28, 2013Published: Sep 1, 2016
Est. expiryNov 28, 2033(~7.3 yrs left)· nominal 20-yr term from priority
Inventors:Pavel S. SmirnovPiotr Semenov
G06V 10/443G06V 10/56G06V 10/462G06K 9/6212G06K 9/6214G06K 9/4652
33
PatentIndex Score
0
Cited by
0
References
0
Claims

Abstract

Technologies for multi-channel feature detection include a computing device to determine a filter response of each image channel of a multi-channel image for one or more image filters. The computing device determines a local differentiating color vector based on the filter responses, applies the filter responses to the local differentiating color vector to generate an adapted response, and determines a total response of the multi-channel image based on the adapted response.

Claims

exact text as granted — not AI-modified
1 - 25 . (canceled) 
     
     
         26 . A computing device for multi-channel feature detection, the computing device comprising:
 an image filtering module to determine a filter response of each image channel of a multi-channel image for one or more image filters;   a local differentiating color module to determine a local differentiating color vector based on the filter responses; and   a response determination module to (i) apply the filter responses to the local differentiating color vector to generate an adapted response and (ii) determine a total response of the multi-channel image based on the adapted response.   
     
     
         27 . The computing device of  claim 26 , wherein to determine the local differentiating color vector comprises to determine a vector that is collinear with a vector of total responses determined for each image channel based on the filter response of each image channel. 
     
     
         28 . The computing device of  claim 26 , wherein to determine the local differentiating color vector comprises to:
 determine a symmetric form of a quadratic form matrix for the multi-channel image; and   identify an eigenvector corresponding with a smallest-valued eigenvalue or a largest-valued eigenvalue of the quadratic form matrix.   
     
     
         29 . The computing device of  claim 28 , wherein to determine the quadratic form matrix comprises to calculate a matrix A={q ij } for image channels i and j, where q ij =½(f i   T ·B·f j +f j   T ·B·f i ), and
 wherein q ij  is the element of the matrix A at the i th  row and j th  column, f is a response vector for an image channel corresponding with an index of f based on the filter response of the image channel, T is a transposition operator, and B is a predefined matrix based on the one or more image filters. 
 
     
     
         30 . The computing device of  claim 28 , wherein to determine the local differentiating color vector comprises to normalize the identified eigenvector to generate the local differentiating color vector. 
     
     
         31 . The computing device of  claim 26 , wherein to determine the filter response of each image channel of the multi-channel image comprises to determine a filter response of each image channel of a multi-channel image for each pixel of the multi-channel image. 
     
     
         32 . The computing device of  claim 26 , wherein to determine the filter response of each image channel of the multi-channel image comprises to generate a response vector for each image channel based on the filter response of each image channel. 
     
     
         33 . The computing device of  claim 32 , wherein to apply the filter responses to the local differentiating color vector comprises to calculate a dot product of the local differentiating color vector and the filter responses. 
     
     
         34 . The computing device of  claim 26 , wherein to determine the local differentiating color vector comprises to determine a normalized local differentiating color vector based on the filter responses. 
     
     
         35 . The computing device of  claim 26 , wherein the response determination module is further to suppress spatial non-extreme responses of the total response of the multi-channel image. 
     
     
         36 . The computing device of  claim 35 , wherein to suppress the spatial non-extreme responses comprises to remove non-interest points from the total response of the multi-channel image, wherein the non-interest points are identified based on a pre-defined threshold value. 
     
     
         37 . The computing device of  claim 26 , further comprising a display module to display an image indicative of the total response on a display of the computing device. 
     
     
         38 . The computing device of  claim 26 , wherein the one or more image filters comprise one or more of a first order derivative image filter or a second order derivative image filter. 
     
     
         39 . One or more machine-readable storage media comprising a plurality of instructions stored thereon that, in response to execution by a computing device, cause the computing device to:
 determine a filter response of each image channel of a multi-channel image for one or more image filters;   determine a local differentiating color vector based on the filter responses;   apply the filter responses to the local differentiating color vector to generate an adapted response; and   determine a total response of the multi-channel image based on the adapted response.   
     
     
         40 . The one or more machine-readable storage media of  claim 39 , wherein to determine the local differentiating color vector comprises to determine a vector that is collinear with a vector of total responses determined for each image channel based on the filter response of each image channel. 
     
     
         41 . The one or more machine-readable storage media of  claim 39 , wherein to determine the local differentiating color vector comprises to:
 determine a symmetric form matrix for the multi-channel image; and   identify an eigenvector corresponding with a smallest-valued eigenvalue or largest-valued eigenvalue of the symmetric form matrix.   
     
     
         42 . The one or more machine-readable storage media of  claim 39 , wherein to determine the local differentiating color vector comprises to:
 determine a symmetric form of a quadratic form matrix for the multi-channel image; and   identify an eigenvector corresponding with a smallest-valued eigenvalue or a largest-valued eigenvalue of the quadratic form matrix.   
     
     
         43 . The one or more machine-readable storage media of  claim 42 , wherein to determine the quadratic form matrix comprises to calculate a matrix A={q ij } for image channels i and j, where q ij =½ (f i   T ·B·f j +f j   T ·B·f i ), and
 wherein q ij  is the element of the matrix A at the i th  row and j th  column, f is a response vector for an image channel corresponding with an index of f based on the filter response of the image channel, T is a transposition operator, and B is a predefined matrix based on the one or more image filters. 
 
     
     
         44 . The one or more machine-readable storage media of  claim 42 , wherein to determine the local differentiating color vector comprises to normalize the identified eigenvector to generate the local differentiating color vector. 
     
     
         45 . The one or more machine-readable storage media of  claim 39 , wherein to determine the filter response of each image channel of the multi-channel image comprises to determine a filter response of each image channel of a multi-channel image for each pixel of the multi-channel image. 
     
     
         46 . The one or more machine-readable storage media of  claim 39 , wherein to apply the filter responses to the local differentiating color vector comprises to calculate a dot product of the local differentiating color vector and the filter responses. 
     
     
         47 . The one or more machine-readable storage media of  claim 39 , wherein the plurality of instructions further causes the computing device to suppress spatial non-extreme responses of the total response of the multi-channel image. 
     
     
         48 . A computing device for multi-channel feature detection, the computing device comprising:
 a local differentiating color module to determine a local differentiating color vector based on pixel values of each image channel of a multi-channel image; and   a response determination module to (i) apply the pixel values of each image channel to the local differentiating color vector to generate an adapted response and (ii) determine a total response of the multi-channel image based on the adapted response.   
     
     
         49 . The computing device of  claim 48 , wherein to determine the local differentiating color vector comprises to determine a vector that is collinear with a vector of total responses determined for each image channel based on the pixel values of each image channel. 
     
     
         50 . The computing device of  claim 48 , wherein to determine the local differentiating color vector comprises to (i) determine a symmetric form matrix for the multi-channel image and (ii) identify an eigenvector corresponding with a smallest-valued eigenvalue or largest-valued eigenvalue of the symmetric form matrix, and
 wherein the response determination module is further to suppress spatial non-extreme responses of the total response of the multi-channel image.

Join the waitlist — get patent alerts

Track US2016253574A1 — get alerts on status changes and closely related new filings.

We store only your email — no account needed. See our privacy policy.