公开(公告)号：US10997740B2

公开(公告)日：2021-05-04

申请号：US16511892

申请日：2019-07-15

Applicant: HERE GLOBAL B.V.

Inventor： Souham Biswas ,  Sanjay Kumar Boddhu

IPC: G06T7/536 ,  G06T7/73 ,  G06K9/00 ,  G06T7/11

Abstract: An approach is provided for estimating a real-world depth information from a monocular image. The approach, for example, involves determining a vanishing point of the monocular image captured by a camera. The approach also involves generating a vanishing point ray from an optical center of the camera through the vanishing point on an image plane of the monocular image to infinity. The approach further involves generating a center line ray from the optical center through a geometric center of the image plane to a feature line that is parallel to the vanishing point ray at a lateral distance. The approach further involves generating a feature ray from the optical center through a location of the feature on the image plane to the feature line. The approach further involves computing the real-world distances of the feature based on image coordinates of the rays, lines, angles derived therefrom, and a known pixel-wise distance of the monocular image.

公开(公告)号：US11600006B2

公开(公告)日：2023-03-07

申请号：US16171814

申请日：2018-10-26

Applicant: HERE Global B.V.

Inventor： Souham Biswas ,  Sanjay Kumar Boddhu

IPC: G06T7/187 ,  G06N5/04 ,  G05D1/02 ,  G06N3/04 ,  G06V10/26 ,  G06N5/046

Abstract: An apparatus and method for encoding objects in a camera-captured image with a deep neural network pipeline including multiple convolutional neural networks or convolutional layers. After identifying at least a portion of the camera-capture image, a first convolutional layer is applied to the at least the portion of the camera-captured image and multiple subregion representations are pooled from the output of the first convolutional layer. One or more additional convolutions are performed. At least one deconvolution is performed and concatenated with the output of one or more convolutions. One or more final convolutions are performed. The at least the portion of the camera-captured image is classified as an object category in response to an output of the one or more final convolutions.

公开(公告)号：US11710239B2

公开(公告)日：2023-07-25

申请号：US17094501

申请日：2020-11-10

Applicant: HERE Global B.V.

Inventor： Souham Biswas ,  Sanjay Kumar Boddhu

IPC: G06T7/00 ,  G06T7/11 ,  G06T7/136 ,  G06T5/20 ,  G06N3/088 ,  G06N3/084

CPC classification number: G06T7/11 ,  G06N3/084 ,  G06N3/088 ,  G06T5/20 ,  G06T7/136 ,  G06T2207/20016 ,  G06T2207/20024 ,  G06T2207/20081 ,  G06T2207/20084

Abstract: An approach is provided for using a machine learning model for identifying planar region(s) in an image. The approach involves, for example, determining the model for performing image segmentation. The model comprises at least: a trainable filter that convolves the image to generate an input volume comprising a projection of the image at different resolution scales; and feature(s) to identify image region(s) having a texture within a similarity threshold. The approach also involves processing the image using the model by generating the input volume from the image using the trainable filter and extracting the feature(s) from the input volume to determine the region(s) having the texture. The approach further involves determining the planar region(s) by clustering the image regions. The approach further involves generating a planar mask based on the planar region(s). The approach further involves providing the planar mask as an output of the image segmentation.