Image recognizing apparatus and image recognizing method

Patent metadata

Field	Value
Publication number	US-2016110606-A1
Application number	US-201514885197-A
Country	US
Kind code	A1
Filing date	Oct 16, 2015
Priority date	Oct 17, 2014
Publication date	Apr 21, 2016
Grant date	—

Abstract

An image recognizing apparatus, including: a camera which outputs at least one of a front image and a rear image of a vehicle; a disturbance estimating unit which outputs an image which is estimated as a disturbance according to a visibility of an image; a disturbance classifying unit which outputs a cause of disturbance according to a brightness of a pixel in the disturbance estimated image; a disturbance result determining unit which determines whether the disturbance is generated in the image in accordance with accumulated cause of the disturbance; and an image recognizing unit which informs the driver that the image is not recognized when it is determined that disturbance occurs in the image.

First claim

What is claimed is:

1 . An image recognizing apparatus, comprising:
a camera configured to output at least one of a front image and a rear image of a vehicle; an disturbance estimating unit configured to output an image which is estimated as a disturbance according to a visibility of an image output from the camera; a disturbance classifying unit configured to output a cause of disturbance according to brightness of pixels in the disturbance estimated image; a disturbance result determining unit configured to determine whether the disturbance is generated in the image in accordance with accumulated cause of the disturbance; and an image recognizing unit configured to inform the driver that the image is not recognized when it is determined that disturbance occurs in the image.

2 . The image recognizing apparatus of claim 1 , wherein the disturbance estimating unit is configured to convert an image output from the camera into a gray image.

3 . The image recognizing apparatus of claim 2 , wherein the disturbance estimating unit sets a region of interest in the gray image.

4 . The image recognizing apparatus of claim 3 , wherein the disturbance estimating unit measures a brightness variation between adjacent pixels of the image in the region of interest of the gray image to output an image which is estimated as disturbance.

5 . The image recognizing apparatus of claim 1 , wherein, when a number of pixels having at least a predetermined brightness level in the set region of the disturbance estimated image is greater than or equal to a predetermined threshold value, the disturbance classifying unit determines that the disturbance estimated image is saturated.

6 . The image recognizing apparatus of claim 1 , wherein, when the number of the pixels having at least a predetermined brightness level set in the disturbance estimated image is less than a predetermined ratio, the disturbance classifying unit determines that the disturbance estimated image has low luminance.

7 . The image recognizing apparatus of claim 1 , wherein, when the disturbance of the disturbance estimated image is determined, the disturbance classifying unit determines the disturbance result of the image by storing whether there is disturbance in a first buffer.

8 . The image recognizing apparatus of claim 1 , wherein:
the disturbance classifying unit is configured to calculate a disturbed light size from the disturbance estimated image to detect fog; the image recognizing apparatus further comprises a foreign substance detecting unit configured to detect illumination of white light in the disturbance estimated image and to compare brightness/color data to detect a foreign substance on a lens of the camera; and the disturbance result determining unit is configured to store the detection number of fog in a first buffer and the detection number of foreign substance in a second buffer, and determine whether either of the detection number of fog or the detection number of foreign substance exceeds a predetermined number to determine whether the disturbance occurs in the image.

9 . The image recognizing apparatus of claim 8 , wherein the disturbance classifying unit is configured to output one result for every frame of the image to the first buffer.

10 . The image recognizing apparatus of claim 8 , wherein the foreign substance detecting unit outputs one result for every five frames of the image to the second buffer.

11 . An image recognizing method, comprising:
outputting at least one of a front image and a rear image of a vehicle; outputting an image which is estimated as a disturbance in according to a visibility of an image; outputting a cause of disturbance according to a brightness of a pixel in the disturbance estimated image; and outputting different results according to the cause of disturbance to a first buffer and a second buffer to accumulatively store the results.

12 . The image recognizing method of claim 11 , wherein, in the outputting of a disturbance estimated image, the image is converted into a gray image and a region of interest is set in the gray image.

13 . The image recognizing method of claim 12 , wherein in the outputting of a disturbance estimated image, a brightness variation between adjacent pixels in the region of interest of the gray image is measured to output an image which is estimated as disturbance.

14 . The image recognizing method of claim 11 , wherein in the outputting of a cause of disturbance, when a number of pixels having at least a predetermined brightness level set in the set region of the disturbance estimated image is greater than or equal to a predetermined threshold value, it is determined that the disturbance estimated image is saturated.

15 . The image recognizing method of claim 11 , wherein in the outputting of a cause of disturbance, when the number of the pixels having at least a predetermined brightness level set in the disturbance estimated image is less than a predetermined value, it is determined that the disturbance estimated image has low luminance.

16 . The image recognizing method of claim 11 , wherein in the accumulatively storing of a cause of disturbance, five results for every frame of the image are stored in the first and second buffers.

17 . The image recognizing method of claim 11 , wherein in the accumulatively storing of a cause of disturbance, the result is stored in the first and second buffers depending on whether a foreign substance in a camera which photographs the image is detected, fog, and image disturbance.

18 . The image recognizing method of claim 17 , wherein, when the number of times of detecting a foreign substance is greater than a number of normal results from the result which is stored in the buffer, it is determined that the foreign substance is present in the image so that the image has low luminance.

19 . The image recognizing method of claim 17 , wherein, when the number of normal result is greater than the number of times of detecting fog and image disturbance from the result which is stored in the buffer, it is determined that the image is saturated.

20 . The image recognizing method of claim 11 , further comprising informing a driver that the image is not recognized when the disturbance is generated in the image.

Assignees

• Hyundai Mobis Co Ltd

Inventors

• Lee Sung Joo

Classifications

• G06V10/776

  Validation; Performance evaluation · CPC title

• G06F18/217

  Validation; Performance evaluation; Active pattern learning techniques · CPC title

• G06V20/588 Primary

  Recognition of the road, e.g. of lane markings; Recognition of the vehicle driving pattern in relation to the road · CPC title

• G06K9/00711 Primary

  Physics · mapped topic

• G06K9/6267

  Physics · mapped topic