METHOD, APPARATUS AND SYSTEM FOR PROCESSING A DISPLAY FROM A SURROUND VIEW CAMERA SOLUTION

• US 20150254825A1
• Filed: 03/09/2015
• Published: 09/10/2015
• Est. Priority Date: 03/07/2014
• Status: Active Grant

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Batur, Aziz Umit, Zhang, Buyue, Pekkucuksen, Ibrahim, Appia, Vikram
• Granted Patent: US 9,533,618 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: B60R 1/00 Optical viewing arrangements; Real-time viewing arrangements for drivers or passengers using optical image capturing systems, e.g. cameras or video systems specially adapted for use in or on vehicles
• CPC Class Codes: B60R 1/27 providing all-round vision, e.g. using omnidirectional cameras
• CPC Class Codes: B60R 2300/303 using joined images, e.g. multiple camera images
• CPC Class Codes: B60R 2300/607 from a bird's eye viewpoint
• CPC Class Codes: G06T 3/4038 Image mosaicing, e.g. composing plane images from plane sub-images

Abstract

A method, apparatus and a system multi-camera image processing method. The method includes performing geometric alignment to produce a geometric output, performing photometric alignment to produce a photometric output and blending output, using data from the geometric alignment and the photometric alignment for performing synthesis function for at least one of blending and stitching images from the multi-cameras, and displaying an image from the synthesis function.

Claims

1. A multi-camera image processing method, comprising:

performing geometric alignment to produce a geometric output;

performing photometric alignment to produce a photometric output and a blending output;

using output from the geometric alignment and the photometric alignment for performing synthesis function for at least one of blending and stitching images from the multi-cameras; and

displaying an image from the synthesis function.

2. The multi-camera image processing method of claim 1, wherein the synthesis function comprises using an alpha blend to blend images from adjacent cameras at a seam boundary to eliminate irregularities at the seam caused by image data from two adjacent cameras.

3. The multi-camera image processing method of claim 1, where in the geometric function comprises a camera calibration.

4. The multi-camera image processing method of claim 3, wherein the camera calibration is at least one of a dynamic calibration and a one-time calibration.

5. The multi-camera image processing method of claim 1, wherein the input to the synthesis function are fisheye frames from the multi-cameras, the geometric output, the photometric output.

6. The multi-camera image processing method of claim 1, wherein the geometric output, the photometric output and blending output are saved in the form of lookup table in memory, a geometric lookup table and a photometric lookup table, respectively.

7. The multi-camera image processing method of claim 1, wherein the output of the photometric lookup table has 255 entries for 8-bit data input.

8. The multi-camera image processing method of claim 1, wherein geometric alignment comprises camera calibration uses calibration charts independent of any camera parameters.

9. The multi-camera image processing method of claim 1, wherein geometric alignment comprises:

performing initial perspective correction on each frame;

running corner detection in the overlapping areas to find regions of interest;

filtering corner data to locate a strong corner;

calculating at least one descriptor of the corner feature and matching the corner from two cameras using scores;

finding perspective matrices for each frame for minimizing distances between matched features; and

creating a geometric lookup table that encodes fisheye lens distortions correction and perspective transformation information.

10. An image processing system, comprising:

a processor;

a memory coupled to the processor;

an image processing module coupled to the processor for executing a multi-camera image processing method, the method comprising;

performing geometric alignment to produce a geometric output;

performing photometric alignment to produce a photometric output and blending output;

using data from the geometric alignment and the photometric alignment for performing synthesis function for at least one of blending and stitching images from the multi-cameras; and

displaying an image from the synthesis function.

11. The image processing system of claim 10, wherein the synthesis function comprises using an alpha blend to blend images from adjacent cameras at a seam boundary to eliminate irregularities at the seam caused by image data from two adjacent cameras.

12. The image processing system of claim 10, where in the geometric function comprises a camera calibration.

13. The image processing system of claim 12, wherein the camera calibration is at least one of a dynamic calibration and a one-time calibration.

14. The image processing system of claim 10, wherein the input to the synthesis function are fisheye frames from the multi-cameras, the geometric output, the photometric output.

15. The image processing system of claim 10, wherein the geometric output, the photometric output and blending output are saved in the form of lookup table in memory, a geometric lookup table and a photometric lookup table, respectively.

16. The image processing system of claim 10, wherein the output of the photometric lookup table has 255 entries for 8-bit data input.

17. The image processing system of claim 10, wherein geometric alignment comprises camera calibration uses calibration charts independent of any camera parameters.

18. The image processing system of claim 10, wherein geometric alignment comprises:

performing initial perspective correction on each frame;

running corner detection in the overlapping areas to find regions of interest;

filtering corner data to locate a strong corner;

calculating at least one descriptor of the corner feature and matching the corner from two cameras using scores;

finding perspective matrices for each frame for minimizing distances between matched features; and

creating a geometric lookup table that encodes fisheye lens distortions correction and perspective transformation information.