Image processing apparatus and its adjustment method

• US 9,992,468 B2
• Filed: 02/14/2012
• Issued: 06/05/2018
• Est. Priority Date: 02/15/2011
• Status: Active Grant

• Current Assignee: DENSO Corporation
• Original Assignee: DENSO Corporation
• Inventors: Osanai, Satoshi, Sato, Hironori
• Primary Examiner(s): Jebari, Mohammed
• Assistant Examiner(s): Vazquez Colon, Maria
• Application Number:
  US13/372,991
• Publication Number:
  US 20120206600A1
• Time in Patent Office: 2,303 Days
• Field of Search: 348148
• CPC Class Codes: G06V 10/56 relating to colour
• CPC Class Codes: H04N 1/6027 Correction or control of colour gradation or colour contrast H04N1/6058 takes precedence
• CPC Class Codes: H04N 1/6047 wherein the test pattern is part of an arbitrary user image
• CPC Class Codes: H04N 9/68 for controlling the amplitude of colour signals, e.g. automatic chroma control circuits H04N9/71, H04N9/73 take precedence; camera processing pipelines for controlling the colour saturation of colour signals H04N23/86

Abstract

In an image processing apparatus, an image acquisition unit acquires an image. A correction unit corrects pixel values of pixels composed of the acquired image by using correction coefficients to produce a correct image. An object recognition unit recognizes an object in the corrected image. A storage unit stores in advance standard color characteristics of a predetermined object which is a recognition target. A color characteristics acquisition unit acquires color characteristics of the predetermined object if the predetermined object is recognized by the object recognition unit. A comparison unit compares the color characteristics of the acquired object with the standard color characteristics of the same predetermined object stored in advance in the storage unit. An updating unit updates the correction coefficients used in the correction unit based on a result of comparison so as to reduce a difference between the color characteristics and the standard color characteristics.

Claims

1. An image processing apparatus, comprising:

an image acquisition unit that acquires an image;

a correction unit that corrects pixel values of the image acquired by the image acquisition unit by using correction coefficients to produce a corrected image, wherein the pixel values include pixel values for red, green, and blue pixels of image data configuring the image, the correction coefficients include a designated correction coefficient for a red pixel, a green pixel, and a blue pixel, the corrected image being configured by image data having corrected pixel values that are calculated by multiplying the pixel values of the image data with respective designated correction coefficients for the red pixel, the green pixel, and the blue pixel, the designated correction coefficient being predetermined and independent of ambient light;

an object recognition unit that recognizes an object in the corrected image;

a storage unit that stores in advance standard color characteristics of a predetermined object that is a recognition target, the recognition target being a predetermined light with a fixed color;

a color characteristics acquisition unit that if the object is recognized by the object recognition unit, acquires color characteristics of the recognized object;

a comparison unit that compares the color characteristics of the recognized object acquired by the color characteristics acquisition unit with the standard color characteristics of the same predetermined object stored in advance by the storage unit; and

an updating unit that updates the designated correction coefficients of the red pixel, the green pixel, and the blue pixel used by the correction unit based on a result of comparison of the color characteristics and the standard color characteristics so as to reduce a difference between the color characteristics and the standard color characteristics, wherein;

the comparison unit compares the color characteristics with the standard color characteristics in a predetermined color space; and

the updating unit acquires an extent of deviation of the color characteristics from the standard color characteristics in the color space, and updates the designated correction coefficients of the red pixel, the green pixel, and the blue pixel based on the extent of deviation in the color space.

3. The image processing apparatus according to claim 1, wherein:

the image acquisition unit is a camera mounted on a vehicle;

the correction unit, the object recognition unit, the storage unit, the color characteristics acquisition unit, the comparison unit, and the updating unit are included in at least one processor mounted on vehicle; and

the processor includes a control unit that controls a state of an on-board device mounted on the vehicle based on the object recognized by the object recognition unit.

7. The image processing apparatus according to claim 1, wherein:

the storage unit stores in advance standard characteristic of a plurality of different predetermined objects, each object being a recognition target, the recognition target being a predetermined light with a fixed color; and

the comparison unit compares the color characteristic of the recognized object with the color characteristic of a matching one of the plurality of different predetermined objects.

8. The image processing apparatus according to claim 7, wherein the plurality of different predetermined objects includes at least one of a vehicle taillight, a traffic light and a road sign.

9. The image processing apparatus according to claim 1, wherein:

the storage unit stores in advance a standard characteristic of a predetermined object, the predetermined object being a recognition target and corresponding to a vehicle taillight, the recognition target having a predetermined light with a fixed color; and

the comparison unit compares the color characteristic of the recognized object with the color characteristic of the predetermined object.

10. The image processing apparatus according to claim 9, wherein:

the storage unit stores in advance a standard characteristic of an additional predetermined object, the additional predetermined object being an additional recognition target and corresponding to at least one of a traffic light and a road sign, the additional recognition target having an additional predetermined light with an additional fixed color; and

the comparison unit compares the color characteristic of the recognized object with the color characteristic of a matching one of the predetermined object and the additional predetermined object.

2. An image processing apparatus comprising:

an image acquisition unit that acquires an image;

a correction unit that corrects pixel values of the image acquired by the image acquisition unit by using correction coefficients to produce a corrected image, wherein the pixel values include pixel values for red, green, and blue pixels of image data configuring the image, the correction coefficients include a designated correction coefficient for a red pixel, a green pixel, and a blue pixel, the corrected image being configured by image data having corrected pixel values that are calculated by multiplying the pixel values of the image data with respective designated correction coefficients for the red pixel, the green pixel, and the blue pixel;

an object recognition unit that recognizes an object in the corrected image;

a storage unit that stores in advance standard color characteristics of a predetermined object that is a recognition target, the recognition target being a predetermined light with a fixed color;

a color characteristics acquisition unit that if the object is recognized by the object recognition unit, acquires color characteristics of the recognized object;

a comparison unit that compares the color characteristics of the recognized object acquired by the color characteristics acquisition unit with the standard color characteristics of the same predetermined object stored in advance by the storage unit; and

an updating unit that updates the designated correction coefficients of the red pixel, the green pixel, and the blue pixel used by the correction unit based on a result of comparison of the color characteristics and the standard color characteristics so as to reduce a difference between the color characteristics and the standard color characteristics, wherein;

the comparison unit compares the color characteristics with the standard color characteristics in a predetermined color space; and

the updating unit gradually varies the designated correction coefficients of the red pixel, the green pixel, and the blue pixel, acquires the color characteristics of the predetermined object using the varied designated correction coefficients of the red pixel, the green pixel, and the blue pixel, compares the color characteristics with the standard color characteristics in the color space, calculates a difference between the color characteristics and the standard color characteristics in the color space, and sets the designated correction coefficients of the red pixel, the green pixel, and the blue pixel that minimizes the difference as updated correction coefficients.

4. A method for adjusting an image processing apparatus that includes an image acquisition unit which acquires an image, a correction unit that corrects pixel values of pixels composed of an acquired image by using correction coefficients, the pixel values including pixel values for red, green, and blue pixels of image data configuring the image, the correction coefficients including a designated correction coefficient for a red pixel, a green pixel, and a blue pixel, a corrected image being configured by image data having corrected pixel values that are calculated by multiplying the pixel values of the image data with respective designated correction coefficients for the red pixel, the green pixel, and the blue pixel, and an object recognition unit that recognizes an object in the corrected image whose pixel values are corrected, the method comprising, the designated correction coefficient being predetermined and independent of ambient light:

storing standard color characteristics of an image of a predetermined standard light source in a storage unit in advance;

acquiring, at the image acquisition unit, an image of the predetermined standard light source;

correcting, at the correction unit, the acquired image of the predetermined standard light source;

comparing, at a comparison unit, color characteristics of the corrected image of the predetermined standard light source with the standard color characteristics of the image of the predetermined standard light source stored in advance in the storage unit; and

updating, at an updating unit, the designated correction coefficients of the red pixel, the green pixel, and the blue pixel based on a result of comparison of the color characteristics of the corrected image of the predetermined standard light source and the standard color characteristics of the image of the predetermined standard light source so as to reduce a difference between the color characteristics of the corrected image of the predetermined standard light source and the standard color characteristics of the image of the predetermined standard light source, wherein;

in the comparing at the comparison unit, the color characteristics are compared with the standard color characteristics in a predetermined color space; and

in the updating at the updating unit, an extent of deviation of the color characteristics from the standard color characteristics in the color space is acquired, and the correction coefficients for red, green, and blue pixels are updated based on the extent of deviation in the color space.

6. The method according to claim 4, wherein:

the image acquisition unit is a camera mounted on a vehicle;

the correction unit, the object recognition unit, the storage unit, the comparison unit, and the updating unit are included in at least one processor mounted on vehicle; and

an on-board device that is mounted on the vehicle is controlled through a control unit of the processor based on the object recognized by the object recognition unit.

5. A method for adjusting an image processing apparatus that includes an image acquisition unit which acquires an image, a correction unit that corrects pixel values of pixels composed of an acquired image by using correction coefficients, the pixel values including pixel values for red, green, and blue pixels of image data configuring the image, the correction coefficients including a designated correction coefficient for a red pixel, a green pixel, and a blue pixel, a corrected image being configured by image data having corrected pixel values that are calculated by multiplying the pixel values of the image data with respective designated correction coefficients for the red pixel, the green pixel, and the blue pixel, and an object recognition unit that recognizes an object in the corrected image whose pixel values are corrected, the method comprising:

storing standard color characteristics of an image of a predetermined standard light source in a storage unit in advance;

acquiring, at the image acquisition unit, an image of the predetermined standard light source;

correcting, at the correction unit, the acquired image of the predetermined standard light source;

comparing, at a comparison unit, color characteristics of the corrected image of the predetermined standard light source with the standard color characteristics of the image of the predetermined standard light source stored in advance in the storage unit; and

updating, at an updating unit, the designated correction coefficients of the red pixel, the green pixel, and the blue pixel based on a result of comparison of the color characteristics of the corrected image of the predetermined standard light source and the standard color characteristics of the image of the predetermined standard light source so as to reduce a difference between the color characteristics of the corrected image of the predetermined standard light source and the standard color characteristics of the image of the predetermined standard light source, wherein;

in the comparing at the comparison unit, the color characteristics are compared with the standard color characteristics in a predetermined color space; and

in the updating at the updating unit, the designated correction coefficients of the red pixel, the green pixel, and the blue pixel are gradually varied, the color characteristics of the predetermined standard light source are acquired using the varied designated correction coefficients of the red pixel, the green pixel, and the blue pixel, the color characteristics are compared with the standard color characteristics in the color space, a difference between the color characteristics and the standard color characteristics in the color space is calculated, and the designated correction coefficients of the red pixel, the green pixel, and the blue pixel that minimize the difference are set as updated correction coefficients.