# US 2022/0012854 A1 · Method and system for image distortion correction using quarter and tangential lookup tables

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0012854 A1
- Document ID: US20220012854A1
- Status: Published application
- Kind: A1
- Application number: 16/923,855
- Filed: 2020-07-08
- Published: 2022-01-13
- Earliest claimed: 2020-07-08
- Assignees: NXP USA, Inc.
- Applicants: NXP USA, Inc.
- Inventors: Ankur Bindal; Michael Andreas Staudenmaier; Sharath Subramanya Naidu
- Web page: https://patentgrep.com/patent/US20220012854A1

## CPC

- G06T 5/80: Geometric correction
- G06T 1/60: Memory management
- H04N 5/91: Television signal processing therefor
- H04N 23/81: for suppressing or minimising disturbance in the image signal generation
- G06T 2207/20021: Dividing image into blocks, subimages or windows
- H04N 5/77: between a recording apparatus and a television camera
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A system and method for correcting image distortion is provided. The system and method remaps pixel position of distorted images using a combination of radial distortion correction and tangential distortion correction lookup tables consuming less physical memory. The solution conserves both memory and memory access bandwidth. The radial distortion correction lookup table is formed by taking advantage of radial distortion being generally symmetric about a determined optical center of the camera lens. This symmetry allows for use of a quarter LUT for correction in all quadrants of a distorted image. In addition, tangential distortion can be corrected in a symmetric manner that saves memory space as well.

## Claims

### Claim 1

A method for populating a corrected image from a distorted image by an image processing circuit, the method comprising:
selecting a portion of the corrected image to populate;
determining a quadrant of the corrected image in which the selected portion resides;
reconstructing a portion of a lookup table (LUT) corresponding to the selected portion of the corrected image, wherein
the LUT correlates pixels in the selected portion of the corrected image with one or more pixels of the distorted image,
said reconstructing is performed using a quarter LUT (QLUT), an x-axis tangential distortion LUT (TxLUT), a y-axis tangential distortion LUT (TyLUT), and the determined quadrant of the corrected image;
populating the selected portion of the corrected image with pixels selected from the distorted image using the generated portion of the LUT; and
selecting a next portion of the corrected image to populate when the corrected image is not fully populated.

### Claim 2 (depends on 1)

The method of claim 1 wherein the QLUT is stored in a memory coupled to the image processing circuit.

### Claim 3 (depends on 1)

The method of claim 1 wherein the QLUT comprises:
a set of values generated from a mean of corresponding values from four quadrants of an original LUT formed around a center point (Cx,Cy) that are flipped such that each quadrant uses the center point as a lowest index.

### Claim 4 (depends on 3)

The method of claim 3 wherein the QLUT comprises an x-axis QLUT (q_x) and a y-axis QLUT (q_y).

### Claim 5 (depends on 3)

The method of claim 3 wherein the original LUT comprises values generated during a calibration that correlates pixels from a corrected calibration image to pixels from a distorted calibration image.

### Claim 6 (depends on 3)

The method of claim 3 wherein the TxLUT comprises a mean of the errors between the original LUT and a reconstructed LUT generated from the QLUT at each x coordinate.

### Claim 7 (depends on 3)

The method of claim 3 wherein the TyLUT comprises a mean of the errors between the original LUT and a reconstructed LUT generated from the QLUT at each y coordinate.

### Claim 8 (depends on 1)

The method of claim 1 further comprising:
selecting a portion of the distorted image corresponding to the selected portion of the corrected image using a predetermined mapping between one or more portions of a corrected image space with corresponding portions of a distorted image space; and
using the selected portion of the distorted image for selecting the pixels from the distorted image in populating the selected portion of the corrected image.

### Claim 9

An image processing system configured to generate a corrected image from a distorted image, the image processing system comprising:
an image memory storing a distorted image and configured to store a corrected image;
a lookup table memory storing a quarter LUT (QLUT), an x-axis tangential distortion LUT (TxLUT), and a y-axis tangential distortion LUT (TyLUT);
an image processing circuit, coupled to the image memory and the lookup table memory, and configured to
select a portion of the corrected image to populate,
determine a quadrant of the corrected image in which the selected portion resides,
generate a portion of a LUT corresponding to the selected portion of the corrected image using the QLUT, TxLUT, TyLUT, and the determined quadrant, wherein the generated LUT correlates pixels in the selected portion of the corrected image with one or more pixels of the distorted image,
populate the selected portion of the corrected image with pixels selected from the distorted image using the generated portion of the LUT, and
select a next portion of the corrected image to populate when the corrected image is not fully populated.

### Claim 10 (depends on 9)

The system of claim 9 wherein the QLUT comprises:
a set of values generated from a mean of corresponding values from four quadrants of an original LUT formed around a center point (Cx,Cy) that are flipped such that each quadrant uses the center point as a lowest index.

### Claim 11 (depends on 10)

The system of claim 10 wherein the QLUT is stored in the lookup table memory as an x-axis QLUT (q_x) and a y-axis QLUT (q_y).

### Claim 12 (depends on 10)

The system of claim 10 wherein the original LUT comprises values generated during a calibration that correlates pixels from a corrected calibration image to pixels from a distorted calibration image.

### Claim 13 (depends on 10)

The system of claim 10 wherein the TxLUT comprises a mean of the errors between the original LUT and a reconstructed LUT generated from the QLUT at each x coordinate.

### Claim 14 (depends on 10)

The system of claim 10 wherein the TyLUT comprises a mean of the errors between the original LUT and a reconstructed LUT generated from the QLUT at each y coordinate.

### Claim 15 (depends on 9)

The system of claim 9 further comprising:
a camera sensor coupled to the image memory and configured to
scan an image, and
store the image in the image memory as the distorted image.

### Claim 16 (depends on 9)

The system of claim 9 further comprising:
mapping memory, coupled to the image processing circuit, and storing a predetermined mapping between one or more portions of a corrected image space with corresponding portions of a distorted image space; and
the image processing circuit is further configured to
select a portion of the distorted image corresponding to the selected portion of the corrected image using the predetermined mapping stored in the mapping memory.

### Claim 17

A method for generating a quarter lookup table (QLUT) from an original lookup table (original LUT) of values correlating a corrected calibration image to pixels from a distorted calibration image, the method comprising:
determining a center point (Cx,Cy) of the original LUT;
breaking the original LUT into four quadrants around (Cx,Cy);
flipping the four quadrants such that each quadrant uses (Cx,Cy) as a lowest index;
generating a mean value of values at common index value for the four quadrants; and
storing the mean value at the index value of the QLUT in a memory of an image processing system.

### Claim 18 (depends on 17)

The method of claim 17 further comprising:
generating the original LUT by mapping each corrected calibration image pixel to a corresponding pixel of the distorted calibration image.

### Claim 19 (depends on 17)

The method of claim 17 wherein said determining the center point of the original LUT comprises:
determining a point of least error for radial and tangential distortion in the original LUT.

### Claim 20 (depends on 17)

The method of claim 17 further comprising performing said method during a calibration of the image processing system, wherein the image processing system comprises a camera having an image sensor and a lens.
