Gaussian image quality analysis tool and method for operation

Patent metadata

Field	Value
Publication number	US-11276156-B2
Application number	US-202016736393-A
Country	US
Kind code	B2
Filing date	Jan 7, 2020
Priority date	Jan 7, 2020
Publication date	Mar 15, 2022
Grant date	Mar 15, 2022

Abstract

A camera image cleaning system of an automobile vehicle includes a camera generating a camera image of a vehicle environment. A processor having a memory executes a control logic to convert the camera image into a grayscale image having multiple image pixels. A convolution equation is retrieved from the memory and is solved to find derivations of the grayscale image defining changes of pixel intensity between consecutive or neighbor ones of the multiple image pixels of the grayscale image. A magnitude and an orientation of the multiple pixels is computed by the processor and used to differentiate weak ones of the image pixels from strong ones of the image pixels.

First claim

What is claimed is:

1. A camera image cleaning system of an automobile vehicle, comprising:
a camera generating a camera image of a vehicle environment;
a processor having a memory and executing a control logic to analyze a camera cleaning system using analyzing steps including:
importing the camera image from the camera system;
converting the camera image to a grayscale image;
finding derivations of the grayscale image using a convolution equation;
computing magnitudes and orientations of image pixels of the grayscale image using a Euclidean distance equation;
differentiating weak ones of the image pixels defined as image pixels that are below a predetermined threshold magnitude and orientation from strong ones of the image pixels;
calculating a variance;
normalizing the variance so that a sharp image having substantially zero blurred ones of the image pixels is assigned a blur number 1 and a blurred image having at least one blurred ones of the image pixels blocked by a vision interference item is assigned a blur number [0,1],
repeating the analyzing steps for a second camera cleaning system; and
comparing the camera cleaning system to the second camera cleaning system and identifying a shortest time to change from an initial blur number for the blurred image to a final blur number for the sharp image.

2. The camera image cleaning system of the automobile vehicle of claim 1 , further including an upper threshold bound and a lower threshold bound applied to further differentiate the strong ones of the image pixels from the weak ones of the image pixels having individual ones of the image pixels with magnitudes greater than the upper threshold bound retained.

3. The camera image cleaning system of the automobile vehicle of claim 2 , further including ones of the image pixels with magnitudes smaller than the lower threshold bound being filtered out.

4. The camera image cleaning system of the automobile vehicle of claim 2 , wherein for the image pixels having magnitudes in-between the upper threshold bound and the lower threshold bound a comparison to eight neighbor pixels is applied.

5. The camera image cleaning system of the automobile vehicle of claim 4 , wherein ones of the image pixels are retained if a magnitude of the ones of the image pixels in-between the upper threshold bound and the lower threshold bound is greater than other image pixels neighboring the ones of the image pixels.

6. The camera image cleaning system of the automobile vehicle of claim 5 , wherein if a magnitude of any of the other image pixels neighboring the ones of the image pixels in between the upper threshold bound and the lower threshold bound is greater than the magnitude of the ones of the image pixels in-between the upper threshold bound and the lower threshold bound the ones of the image pixels in between the upper threshold bound and the lower threshold bound are filtered out.

7. The camera image cleaning system of the automobile vehicle of claim 1 , wherein the image pixels of the grayscale image include white pixels and black pixels, the processor identifying if the grayscale image has more of the white pixels than the black pixels during the calculation of the variance.

8. The camera image cleaning system of the automobile vehicle of claim 1 , further including an 8-connected method used to differentiate the weak ones of the image pixels from the strong ones of the image pixels.

9. A method for operating a gaussian image quality analysis system of an automobile vehicle, comprising:
analyzing a camera cleaning system using analyzing steps including:
importing a camera image from a camera system;
converting the camera image to a grayscale image;
finding derivations of the grayscale image using a convolution equation;
computing magnitudes and orientations of image pixels of the grayscale image using a Euclidean distance equation;
differentiating weak ones of the image pixels defined as image pixels that are below a predetermined threshold magnitude and orientation from strong ones of the image pixels;
calculating a variance;
normalizing the variance so that a sharp image having substantially zero blurred ones of the image pixels is assigned a blur number 1 and a blurred image having at least one blurred ones of the image pixels blocked by a vision interference item is assigned a blur number [0,1],
repeating the analyzing steps for a second camera cleaning system; and
comparing the camera cleaning system to the second camera cleaning system and identifying a shortest time to change from an initial blur number for the blurred image to a final blur number for the sharp image.

10. The method for operating the gaussian image quality analysis system of the automobile vehicle of claim 9 , further including choosing a fixed time interval after the camera cleaning system is initiated and the second camera cleaning system is initiated.

11. The method for operating the gaussian image quality analysis system of the automobile vehicle of claim 10 , further including measuring blur number increases for the camera cleaning system and the second camera cleaning system over the fixed time interval.

12. The method for operating the gaussian image quality analysis system of the automobile vehicle of claim 9 , wherein the importing the camera image from the camera system omits a previous camera image to eliminate a vibrational input impacting the previous camera image.

13. The method for operating the gaussian image quality analysis system of the automobile vehicle of claim 9 , further including:
defining an upper threshold bound and a lower threshold bound for the image pixels;
filtering out image pixels with magnitudes smaller than the lower threshold bound; and
retaining image pixels with magnitudes greater than the upper threshold bound.

14. A method for operating a gaussian image quality analysis system of an automobile vehicle, comprising:
analyzing a first camera cleaning system using analyzing steps including:
importing multiple camera images from a camera system;
converting the camera images to grayscale images;
finding derivations of the grayscale images as sequential frames using a convolution equation;
computing magnitudes and orientations of image pixels of the grayscale images using a Euclidean distance equation;
differentiating weak ones of the image pixels defined as image pixels that are below a predetermined threshold magnitude and orientation from strong ones of the image pixels for the sequential frames; and
calculating a variance for individual ones of the sequential frames;
normalizing the variance so that a sharp image having substantially zero blurred ones of the image pixels is assigned a first blur number and a blurred image having at least one blurred ones of the image pixels blocked by a vision interference item is assigned a second blur number;

repeating the analyzing steps for a second camera cleaning system; and
comparing results of the analyzing steps for the first camera cleaning system to the analyzing steps for the second camera cleaning system and identifying a shortest time to change from an initial blur number for the blurred image to a final blur number for the sharp image.

15. The method for operating the gaussian image quality analysis system of the automobile vehicle of claim 14 , further including filtering out the weak ones of the image pixels and retaining the strong ones of the image pixels defined as image pixels that are above the predetermined threshold magnitude.

16. The method for operating the gaussian image quality analysis system of the

Assignees

• Gm Global Tech Operations Llc

Inventors

• Wang Keyu
• Yadav Sanket

Classifications

• G06T2207/30252

  Vehicle exterior; Vicinity of vehicle · CPC title

• G06T2207/30168

  Image quality inspection · CPC title

• G06T7/90

  Determination of colour characteristics · CPC title

• G06T2207/10016

  Video; Image sequence · CPC title

• G06T7/0002 Primary

  Inspection of images, e.g. flaw detection · CPC title