# US 2019/0171905 A1 · Method, system and apparatus for comparing objects in images > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2019/0171905 A1 - Document ID: US20190171905A1 - Status: Published application - Kind: A1 - Application number: 15/832,288 - Filed: 2017-12-05 - Published: 2019-06-06 - Earliest claimed: 2017-12-05 - Assignees: Canon Kabushiki Kaisha - Applicants: Canon Kabushiki Kaisha - Inventors: Getian Ye; Ka Ming Leung - Web page: https://patentgrep.com/patent/US20190171905A1 ## CPC - G06F 16/7837: {using objects detected or recognised in the video content} - G06F 16/7854: {using shape (G06F16/7837 takes precedence)} - G06F 18/24323: {Tree-organised classifiers} - G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69) - G06V 20/64: Three-dimensional [3D] objects - G06V 40/103: {Static body considered as a whole, e.g. static pedestrian or occupant recognition} - G06V 40/161: {Detection; Localisation; Normalisation} - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract A method of comparing objects in images. A dictionary determined from a plurality of feature vectors formed from a test image and codes formed by applying the dictionary to the feature vectors, is received. The dictionary is based on a modified manifold obtained by determining correspondences for codes using pairwise similarities between codes. Comparison codes are determined for the objects in the images by applying the dictionary to feature vectors of the objects in the images. The objects in the images are compared based on the comparison codes of the objects. ## Claims ### Claim 1 A method of comparing objects in images, the method comprising: receiving a dictionary determined from a plurality of feature vectors formed from a test image and codes formed by applying the dictionary to the feature vectors, wherein the dictionary is based on a modified manifold obtained by determining correspondences for codes using pairwise similarities between codes; determining comparison codes for the objects in the images by applying the dictionary to feature vectors of the objects in the images; and comparing the objects in the images based on the comparison codes of the objects. ### Claim 2 (depends on 1) The method according to claim 1, further comprising: refining the modified manifold based on local connectivity of codes; and determining the dictionary using the refined manifold and codes. ### Claim 3 (depends on 1) The method according to claim 1, further comprising: generating an affinity matrix using the comparison codes; determining a Laplacian matrix using the affinity matrix; and determining the dictionary using the plurality of feature vectors and the Laplacian matrix. ### Claim 4 (depends on 1) The method according to claim 1, further comprising determining a plurality of feature channels for one or more of the images. ### Claim 5 (depends on 1) The method according to claim 1, further comprising determining the dictionary based on a minimized objective function comprised of a reconstruction error. ### Claim 6 (depends on 1) The method according to claim 1, further comprising determining a similarity score between the codes. ### Claim 7 (depends on 6) The method according to claim 6, wherein the similarity score is a Cosine distance between the codes. ### Claim 8 (depends on 6) The method according to claim 6, further comprising comparing the similarity score to a fixed threshold. ### Claim 9 (depends on 6) The method according to claim 6, further comprising comparing the similarity score to a further similarity score for a query object and all other objects in a video stream. ### Claim 10 (depends on 1) The method according to claim 1, further comprising determining a dissimilarity score between the codes. ### Claim 11 (depends on 10) The method according to claim 10, wherein the dissimilarity score is a Euclidean distance between the codes. ### Claim 12 (depends on 1) The method according to claim 1, wherein the feature vectors are based on properties of pixels in the test image. ### Claim 13 An apparatus for comparing objects in images, the apparatus comprising: means for receiving a dictionary determined from a plurality of feature vectors formed from a test image and codes formed by applying the dictionary to the feature vectors, wherein the dictionary is based on a modified manifold obtained by determining correspondences for codes using pairwise similarities between codes; means for determining comparison codes for the objects in the images by applying the dictionary to feature vectors of the objects in the images; and means for comparing the objects in the images based on the comparison codes of the objects. ### Claim 14 A system for comparing objects in images, the system comprising: a memory for storing data and a computer program; a processor coupled to the memory for executing the computer program, the computer program having instructions for: receiving a dictionary determined from a plurality of feature vectors formed from a test image and codes formed by applying the dictionary to the feature vectors, wherein the dictionary is based on a modified manifold obtained by determining correspondences for codes using pairwise similarities between codes, determining comparison codes for the objects in the images by applying the dictionary to feature vectors of the objects in the images, and comparing the objects in the images based on the comparison codes of the objects. ### Claim 15 A non-transitory computer readable medium storing a computer program to cause an apparatus to perform a method of comparing objects in images, the method comprising: receiving a dictionary determined from a plurality of feature vectors formed from a test image and codes formed by applying the dictionary to the feature vectors, wherein the dictionary is based on a modified manifold obtained by determining correspondences for codes using pairwise similarities between codes; determining comparison codes for the objects in the images by applying the dictionary to feature vectors of the objects in the images; and comparing the objects in the images based on the comparison codes of the objects.