US 12,437,025 B2
Two-way descriptor matching on deep learning accelerator
Deepak Kumar Poddar, Bengaluru (IN); Soyeb Nagori, Bengaluru (IN); Hrushikesh Tukaram Garud, Parbhani (IN); and Pramod Kumar Swami, Bengaluru (IN)
Assigned to TEXAS INSTRUMENTS INCORPORATED, Dallas, TX (US)
Filed by Texas Instruments Incorporated, Dallas, TX (US)
Filed on Nov. 1, 2023, as Appl. No. 18/499,627.
Application 18/499,627 is a continuation of application No. 17/149,474, filed on Jan. 14, 2021, granted, now 11,847,184.
Claims priority of application No. 202041001570 (IN), filed on Jan. 14, 2020.
Prior Publication US 2024/0078284 A1, Mar. 7, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 17/16 (2006.01); G06F 7/523 (2006.01); G06F 9/50 (2006.01); G06F 18/22 (2023.01); G06V 10/75 (2022.01)
CPC G06F 17/16 (2013.01) [G06F 7/523 (2013.01); G06F 9/5027 (2013.01); G06F 18/22 (2023.01); G06V 10/75 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
determining a first set of keypoints from a first image and a second set of keypoints from a second image;
obtaining a first descriptor matrix based on the first set of keypoints;
obtaining a second descriptor matrix based on the second set of keypoints;
determining, by a hardware accelerator, a cost matrix based on the first descriptor matrix and the second descriptor matrix;
determining, by the hardware accelerator, a reduced cost matrix; and
identifying matching keypoints based on the reduced cost matrix.