Systems and methods for downsampling images

Patent metadata

Field	Value
Publication number	US-12470836-B2
Application number	US-202218147588-A
Country	US
Kind code	B2
Filing date	Dec 28, 2022
Priority date	Dec 28, 2022
Publication date	Nov 11, 2025
Grant date	Nov 11, 2025

Abstract

Systems and methods are provided of downsampling an image to a plurality of image resolutions. The method comprises capturing, using a camera, an image depicting an environment within view of the camera, identifying a first section of the image depicting an area of the environment spaced within a first distance range from the camera, identifying a second section of the image depicting an area of the environment spaced within a second distance range from the camera, identifying a third section of the image depicting an area of the environment spaced within a third distance range from the camera, and downsampling the first section of the image to a first image resolution, generating a first processed image, the second section of the image to a second image resolution, generating a second processed image, and the third section of the image to a third image resolution, generating a third processed image.

First claim

The invention claimed is:

1 . A method of downsampling an image to a plurality of image resolutions, comprising:
capturing, using a camera, an image depicting an environment within view of the camera; identifying a first section of the image, wherein the first section depicts an area of the environment spaced within a first distance range from the camera; identifying a second section of the image, wherein the second section depicts an area of the environment spaced within a second distance range from the camera; identifying a third section of the image, wherein the third section depicts an area of the environment spaced within a third distance range from the camera; determining a first image resolution for a first processed image, a second image resolution for a second processed image, and a third image resolution for a third processed image based on (i) an angle of the camera relative to a level plane, (ii) a difference in elevations between the camera and each of the first, second or third distance range, and/or (iii) an incline of a road shown in the image, wherein the first distance range, the second distance range, and the third distance range pertain to distance ranges of a driving surface; and downsampling:
the first section of the image to the first image resolution, generating the first processed image;
the second section of the image to the second image resolution, generating the second processed image; and
the third section of the image to the third image resolution, generating the third processed image,
wherein:
the first distance range is farther from the camera than the second distance range, and
the first image resolution is less than the second image resolution, and the second image resolution being different than the third image resolution.

2 . The method of claim 1 , wherein the first distance range, the second distance range, and the third distance range do not overlap.

3 . The method of claim 1 , wherein the second distance range is:
closer to the camera than the first distance range; and farther from the camera than the third distance range.

4 . The method of claim 1 , further comprising generating a distance map of the environment of the image.

5 . The method of claim 4 , wherein the generating the distance map is performed using one or more of the following:
LiDAR; RADAR; and image analysis.

6 . The method of claim 4 , wherein one or more of the following are based on the distance map:
the identifying the first section; the identifying the second section; and the identifying the third section.

7 . The method of claim 1 , wherein:
the downsampling is performed using a processor, and the camera comprises the processor.

8 . The method of claim 1 , further comprising, prior to downsampling the first section of the image, the second section of the image, and the third section of the image:
downsampling the image to an image resolution lower than an original image resolution.

9 . The method of claim 1 , wherein the second image resolution is greater than:
the first image resolution; and the third image resolution.

10 . The method of claim 1 , wherein:
the first distance range is farther from the one or more cameras than the second distance range and the third distance range, and the first image resolution is less than the second image resolution and the third image resolution.

11 . The method of claim 1 , wherein
the first processed image comprises a center cropped version of the image, the second processed image comprises a vertical center-cropped version of the image, and the third processed image comprises an uncropped version of the image.

12 . A system for downsampling an image to a plurality of image resolutions, comprising:
a vehicle; and an imaging module, coupled to the vehicle, the imaging module comprising:
one or more cameras, configured to capture an image depicting an environment within view of the one or more cameras; and
a processor, configured to:
identify:
a first section of the image depicting an area of the environment spaced within a first distance range from the one or more cameras;
a second section of the image depicting an area of the environment spaced within a second distance range from the one or more cameras; and
a third section of the image depicting an area of the environment spaced within a third distance range from the one or more cameras;

determine a first image resolution for a first processed image, a second image resolution for a second processed image, and a third image resolution for a third processed image based on (i) an angle of the camera relative to a level plane, (ii) a difference in elevations between the camera and each of the first, second or third distance range, and/or (iii) an incline of a road shown in the image, wherein the first distance range, the second distance range, and the third distance range pertain to distance ranges of a driving surface; and

downsample:
the first section of the image to the first image resolution, generating the first processed image;
the second section of the image to the second image resolution, generating the second processed image; and
the third section of the image to the third image resolution, generating the third processed image,

wherein:
the first distance range is farther from the one or more cameras than the second distance range, and
the first image resolution is less than the second image resolution.

13 . The system of claim 12 , wherein the first distance range, the second distance range, and the third distance range do not overlap.

14 . The system of claim 12 , wherein the second distance range is:
closer to the one or more cameras than the first distance range; and farther from the one or more cameras than the third distance range.

15 . The system of claim 12 , wherein the processor is further configured to generate a distance map of the environment of the image.

16 . The system of claim 15 , wherein the generating the distance map is performed using one or more of the following:
LiDAR; RADAR; and image analysis.

17 . The system of claim 15 , wherein one or more of the following are based on the distance map:
the identifying the first section; the identifying the second section; and
the identifying the third section.

18 . The system of claim 12 , wherein the one or more cameras comprise the processor.

19 . The system of claim 12 , wherein the processor is further configured to, prior to downsampling the first section of the image, the second section of the image, and the third section of the image:
downsample the image to an image resolution lower than an original image resolution.

20 . The system of claim 12 , wherein the second image resolution is greater than:
the first image resolution; and the third image resolution.

21 . A system, comprising:
an imaging device including one or more cameras, the imaging device coupled to a vehicle,
wherein the one or more cameras are configured to capture an image depicting an environment within view of the one or more cameras; and
a computing device, including a processor and a memory, coupled to the vehicle, configured to store programming instructions that, when executed by the processor, cause the processor to:
identify:
a first section of the image depicting an area of the environment spaced within a first distance range from the one or more cameras;
a second section of the image

Assignees

• Kodiak Robotics Inc

Inventors

• Wendel Andreas

Classifications

• G06T3/40

  Scaling of whole images or parts thereof, e.g. expanding or contracting · CPC title

• H04N13/25

  using two or more image sensors with different characteristics other than in their location or field of view, e.g. having different resolutions or colour pickup characteristics; using image signals from one sensor to control the characteristics of another sensor · CPC title

• G06T2207/30261

  Obstacle · CPC title

• G06T7/50

  Depth or shape recovery · CPC title

• G06V20/58

  Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads · CPC title