IP Library Granted Patent US 10,043,262
Granted Patent B2
US 10,043,262 · App. 15/221,591 · Granted Aug 7, 2018

Abnormal image detection device, image processing system having abnormal image detection device, and vehicle incorporating image processing system

Inventors: Takashi Akasaka (Iwata, JP); Takahiro Ishii (Iwata, JP)
Assignee: YAMAHA HATSUDOKI KABUSHIKI KAISHA
G06T7/0008B60R1/00G06K9/00805G06T7/593H04N13/0203B60R2300/307G06T2207/10021G06T2207/30261G08G1/16
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,043,262
App. No.
15/221,591
Granted
Aug 7, 2018
Kind
B2
Abstract

An abnormal image detection device includes a reliability determination part that determines whether or not each pixel of a parallax image has a low reliability. The reliability determination part has an edge intensity determination section. For each respective pixel of original images, the edge intensity determination section determines whether or not the respective pixel is smaller than or equal to a threshold value, and if the respective pixel is smaller than or equal to the threshold value, the reliability determination part determines that a corresponding one of the pixels in the parallax image, which corresponds to the respective pixel, is a pixel having the low reliability. A computation part computes a total number of pixels of the parallax image that each has the low reliability. An abnormality determination part that when the total number of pixels exceeds a predetermined value, determines that the parallax image is abnormal.

Claims (63)

1. An abnormal image detection device that detects whether or not at least one of a plurality of original images is abnormal, the plurality of original images being obtained from an image pickup detector and used to prepare a parallax image, the parallax image having a plurality of pixels, the abnormal image detection device comprising:

a reliability determination part that determines whether or not each of the pixels of the parallax image has a low reliability, the reliability determination part having

an edge intensity determination section that for each respective pixel of the plurality of original images,

determines whether or not the edge intensity of the respective pixel is smaller than or equal to a threshold value, and

if the edge intensity of the respective pixel is smaller than or equal to the threshold value, determines that one of the pixels in the parallax image, which corresponds to the respective pixel, has the low reliability, and

a parallax value determination section that

determines a parallax value of each of the pixels of the parallax image,

determines whether each parallax value is out of a predetermined range, and

in each instance that a respective one of the parallax values is out of the predetermined range, determines that the pixel of the parallax image, which corresponds to the respective one of the parallax values, has the low reliability;

a computation part that computes a total number of pixels which each has the low reliability as determined by the reliability determination part; and

an abnormality determination part that determines that the parallax image is abnormal, when the total number of pixels computed by the computation part exceeds a predetermined value.

2. The abnormal image detection device according to claim 1 , wherein the parallax image includes at least a first parallax image (Ab) prepared based upon a first original image (A) of the plurality of original images, and a second parallax image (Ba) prepared based upon a second original image (B) of the plurality of original images, and

wherein the reliability determination part has a parallax value difference determination section that determines when a difference (d 1 -d 2 ) is out of a predetermined range, that a first pixel (pA) of the first parallax image (Ab) has low reliability,

wherein d 1 is a parallax value of the first pixel (pA) of the first parallax image (Ab),

wherein d 2 is a parallax value of a second pixel (pB) of the second parallax image (Ba), and

further wherein the second pixel (pB) corresponds to the first pixel (pA).

3. An image processing system comprising:

the abnormal image detection device according to claim 1 ;

the image pickup detector, the image pickup detector including a plurality of image pickup parts obtaining the plurality of original images;

a parallax image preparation part that prepares the parallax image on the basis of the plurality of original images;

a distance computation part that computes a depth distance on the basis of the parallax image prepared by the parallax image preparation part; and

an output part that outputs abnormality information when the abnormal image detection device detects an abnormality in the original images.

4. The image processing system according to claim 3 , wherein the distance computation part is configured to compute a depth distance on the basis of the parallax image other than the parallax image in which an abnormality has been detected.

5. The image processing system according to claim 3 ,

wherein in the parallax image prepared by the parallax image preparation part, the distance computation part is configured to compute a depth distance on the basis of the parallax image formed by removing pixels having low reliability.

6. The vehicle according to claim 5 , further comprising a front windshield in front of the image pickup parts in an image pickup direction in which the original images are obtained by the image pickup parts,

wherein the vehicle is free of any windshield wiper on the front windshield.

7. A vehicle incorporating the image processing system according to claim 3 .

8. An abnormal image detection device that detects whether or not at least one of a plurality of original images is abnormal, the plurality of original images being obtained from an image pickup detector and used to prepare a parallax image, the abnormal image detection device comprising:

a reliability determination part including

an edge intensity determination section that determines whether or not each respective pixel of the plurality of original images is of low reliability by classifying the respective pixel as being of low reliability when an edge intensity of the respective pixel is smaller than or equal to a threshold value, and

a parallax value determination section that

determines a parallax value of each of the pixels of the parallax image,

determines whether each parallax value is out of a predetermined range, and

in each instance that a respective one of the parallax values is out of the predetermined range, determines that the pixel of the parallax image, which corresponds to the respective one of the parallax values, has the low reliability;

a computation part that computes a total number of pixels which each has the low reliability as determined by the reliability determination part; and

an abnormality determination part that determines that the parallax image is abnormal when the total number of pixels computed by the computation part exceeds a predetermined value.

9. An abnormal image detection device that detects whether or not at least one of a plurality of original images is abnormal, the plurality of original images being obtained from an image pickup detector and used to prepare a parallax image, the parallax image having a plurality of pixels, the abnormal image detection device comprising:

a reliability determination part that determines whether or not each of the pixels of the parallax image has a low reliability, the reliability determination part having an edge intensity determination section that for each respective pixel of the original images,

determines whether or not the edge intensity of the respective pixel is smaller than or equal to a threshold value, and

if the edge intensity of the respective pixel is smaller than or equal to the threshold value, determines that one of the pixels in the parallax image, which corresponds to the respective pixel, has the low reliability;

a computation part that computes a total number of pixels which each has the low reliability as determined by the reliability determination part; and

an abnormality determination part that determines that the parallax image is abnormal, when the total number of pixels computed by the computation part exceeds a predetermined value,

wherein the parallax image includes at least a first parallax image (Ab) prepared based upon a first original image (A) of the plurality of original images, and a second parallax image (Ba) prepared based upon a second original image (B) of the plurality of original images, and

wherein the reliability determination part has a parallax value difference determination section that determines when a difference (d 1 -d 2 ) is out of a predetermined range, that a first pixel (pA) of the first parallax image (Ab) has low reliability,

wherein d 1 is a parallax value of the first pixel (pA) of the first parallax image (Ab),

wherein d 2 is a parallax value of a second pixel (pB) of the second parallax image (Ba), and

further wherein the second pixel (pB) corresponds to the first pixel (pA).

10. The abnormal image detection device according to claim 9 , wherein the reliability determination part has a parallax value determination section that

determines a parallax value of each of the pixels of the parallax image,

determines whether each parallax value is out of a predetermined range, and

in each instance that a respective one of the parallax values is out of the predetermined range, determines that the pixel of the parallax image, which corresponds to the respective one of the parallax values, has the low reliability.

11. An image processing system comprising:

the abnormal image detection device according to claim 9 ;

the image pickup detector, the image pickup detector including a plurality of image pickup parts obtaining the plurality of original images;

a parallax image preparation part that prepares the parallax image on the basis of the plurality of original images;

a distance computation part that computes a depth distance on the basis of the parallax image prepared by the parallax image preparation part; and

an output part that outputs abnormality information when the abnormal image detection device detects an abnormality in the original images.

12. The image processing system according to claim 11 , wherein the distance computation part is configured to compute a depth distance on the basis of the parallax image other than the parallax image in which an abnormality has been detected.

13. The image processing system according to claim 11 , wherein in the parallax image prepared by the parallax image preparation part, the distance computation part is configured to compute a depth distance on the basis of the parallax image formed by removing pixels having low reliability.

14. The vehicle according to claim 13 , further comprising a front windshield in front of the image pickup parts in an image pickup direction in which the original images are obtained by the image pickup parts,

wherein the vehicle is free of any windshield wiper on the front windshield.

15. A vehicle incorporating the image processing system according to claim 11 .

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jul 27, 2016
From: AKASAKA, TAKASHI; ISHII, TAKAHIRO
To: YAMAHA HATSUDOKI KABUSHIKI KAISHA
Reel/Frame 039275/0815 →
Priority Claims (5)
JP 2015-149524 · Jul 29, 2015 · national
JP 2015-157183 · Aug 7, 2015 · national
JP 2015-175792 · Sep 7, 2015 · national
JP 2016-043199 · Mar 7, 2016 · national
JP 2016-127534 · Jun 28, 2016 · national
Continuity (1)
Related Publication 20170032517A1 · Feb 2, 2017