-
公开(公告)号：US20230206604A1
公开(公告)日：2023-06-29
申请号：US18176699
申请日：2023-03-01
Applicant: TEXAS INSTRUMENTS INCORPORATED
Inventor： DEEPAK KUMAR PODDAR , SOYEB NAGORI , HRUSHIKESH TUKARAM GARUD , KUMAR DESAPPAN
CPC classification number: G06V10/7715 , G06T3/4046 , G06V10/48 , G06V10/462 , G06N3/084
Abstract: An example image feature extraction system comprises an encoder neural network having a first set of layers and a decoder neural network having a second set of layers and a third set of layers. The encoder neural network receives an input image, processes the input image through the first set of layers, and computes an encoded feature map based on the input image. The decoder neural network receives the encoded feature map, processes the encoded feature map through the second set of layers to compute a keypoint score map, and processes the encoded feature map through at least a portion of the third set of layers to compute a feature description map.
-
公开(公告)号：US20220180476A1
公开(公告)日：2022-06-09
申请号：US17112096
申请日：2020-12-04
Applicant: TEXAS INSTRUMENTS INCORPORATED
Inventor： DEEPAK KUMAR PODDAR , SOYEB NAGORI , HRUSHIKESH TUKARAM GARUD , KUMAR DESAPPAN
Abstract: This description relates to image feature extraction. In some examples, a system can include a keypoint detector and a feature list generator. The keypoint detector can be configured to upsample a keypoint score map to produce an upsampled keypoint score map. The keypoint score map can include feature scores indicative of a likelihood of at least one feature being present at keypoints in an image. The feature list generator can be configured to identify a subset of keypoints of the keypoints in the image using the feature scores of the up sampled keypoint score map, determine descriptors for the subset of keypoints based on a feature description map, and generate a keypoint descriptor map for the image based on the determined descriptors.
-
公开(公告)号：US20170193311A1
公开(公告)日：2017-07-06
申请号：US15346491
申请日：2016-11-08
Applicant: TEXAS INSTRUMENTS INCORPORATED
Inventor： DEEPAK KUMAR PODDAR , SHYAM JAGANNATHAN , SOYEB NAGORI , PRAMOD KUMAR SWAMI
CPC classification number: G06K9/00791 , G05D1/0253 , G05D2201/0213 , G06F9/30007 , G06F15/8007 , G06K9/52 , G06T3/4038
Abstract: A vehicular structure from motion (SfM) system can include an input to receive a sequence of image frames acquired from a camera on a vehicle and an SIMD processor to process 2D feature point input data extracted from the image frames so as to compute 3D points. For a given 3D point, the SfM system can calculate partial ATA and partial ATb matrices outside of an iterative triangulation loop, reducing computational complexity inside the loop. Multiple tracks can be processed together to take full advantage of SIMD instruction parallelism.
-
公开(公告)号：US20170186169A1
公开(公告)日：2017-06-29
申请号：US15235516
申请日：2016-08-12
Applicant: TEXAS INSTRUMENTS INCORPORATED
Inventor： PRASHANTH RAMANATHPU VISWANATH , SOYEB NAGORI , MANU MATHEW
CPC classification number: G06K9/00791 , G06K9/4604 , G06T7/269 , G06T7/579 , G06T2207/30244 , G06T2207/30252
Abstract: A vehicular structure from motion (SfM) system can store a number of image frames acquired from a vehicle-mounted camera in a frame stack according to a frame stack update logic. The SfM system can detect feature points, generate flow tracks, and compute depth values based on the image frames, the depth values to aid control of the vehicle. The frame stack update logic can select a frame to discard from the stack when a new frame is added to the stack, and can be changed from a first in, first out (FIFO) logic to last in, first out (LIFO) logic upon a determination that the vehicle is stationary. An optical flow tracks logic can also be modified based on the determination. The determination can be made based on a dual threshold comparison to insure robust SfM system performance.
-
-
-
Search Results
4
records
(took 0.014 seconds)
