Detection of and correction for specular reflections in digital image acquisition

• US 6,061,091 A
• Filed: 05/22/1997
• Issued: 05/09/2000
• Est. Priority Date: 05/30/1996
• Status: Expired due to Fees

• Current Assignee: Agfa-Gevaert N.V.
• Original Assignee: Agfa-Gevaert N.V.
• Inventors: Geybels, Nick, Van de Poel, Koen, Stokes, Earle
• Primary Examiner(s): GARBER, WENDY RAE
• Assistant Examiner(s): Ferguson, Eric
• Application Number:
  US08/861,929
• Time in Patent Office: 1,083 Days
• Field of Search: 378/98.7, 378/98.8, 378/98.11, 358/522, 348/254, 348/364, 348/241, 382/170, 382/171
• US Class Current: 348/241
• CPC Class Codes: G06T 5/40 using histogram techniques
• CPC Class Codes: G06T 5/92 based on global image properties

Abstract

Image processing parameters for gradation correction of a digital image from a scene, which includes specular reflections may be seriously influenced by the presence of those specular reflections. Therefore, a method is developed to detect the presence of specular reflections in a scene, to modify the exposure time of the photo-sensitive means, which converts a luminous image to electronic image signals and to apply a gradation correction to the image data thus acquired. In one embodiment a digital camera is used to capture the image data, a cumulative histogram is built and two characteristic cumulative frequencies are used to determine via the histogram specular densities and highlight densities. The relative position of these density values indicates whether specular reflections are present, and give suitable parameters for setting an optimal exposure time and for computing a gradation correction curve.

Claims

1. A method for detecting and processing a specular image density level X₁ and a highlight image density level X₂ in an illuminated scene comprising the following steps:

exposing a photo-sensitive means by a luminous image from said scene, thereby generating a plurality of digital image signals, each said digital image signal representing a density of a portion of said luminous image;

establishing a cumulative histogram for the image density levels of said digital image signals;

providing a reference specular cumulative frequency Y₁ ;

providing a reference highlight cumulative frequency Y₂ ;

converting said specular cumulative frequency Y₁ via said cumulative histogram, to obtain said specular image density level X₁ ;

converting said highlight cumulative frequency Y₂ via said cumulative histogram, to obtain said highlight image density level X₂ ; and

indicating a presence of specular reflections when a ratio α

=∥

X_MAX -X₂ ∥

/∥

X_MAX -X₁ ∥

is smaller than a predetermined value, wherein X_MAX is an image density level greater than or equal to a highest density level within said luminous image.

2. The method according to claim 1, wherein said predetermined value is 0.875.

3. The method according to claim 1, wherein:

the specular cumulative frequency Y₁ and

the highlight cumulative frequency Y₂ are selected such that;

the value D₁ =∥

Y₁ -Y_MIN ∥

; and

the value D₂ =∥

Y₂ -Y_MIN ∥

;

wherein Y_MIN is the cumulative frequency of the image level X_MIN corresponding to the lowest density level in the luminous image, and wherein

10^-5 ≦

D₁,

D₁ ≦

(D₂ /2), and

D₂ ≦

0.1.

5. The method according to claim 1, further comprising optimizing an exposure time for exposing said photo-sensitive means based on the image density level of X₁, X₂ or a combination thereof.

7. The method according to claim 1, wherein said image signals are further transformed by a gradation correction curve.

8. The method according to claim 7, wherein said gradation correction curve comprises linear segments, having a start point at X₁, a break point at X₂ and an end point at X_MAX.

9. The method according to claim 8, wherein said gradation correction curve comprises the break point is smoothed by an arc.

10. The device according to claim 1, wherein:

the specular cumulative frequency Y₁ and

the highlight cumulative frequency Y₂ are selected such that;

the value D₁ =∥

Y₁ -Y_MIN ∥

; and

the value D₂ =∥

Y₂ -Y_MIN ∥

;

wherein Y_MIN is the cumulative frequency of the image level X_MIN corresponding to the lowest density level in the luminous image, and wherein

10^-5 ≦

D₁,

D₁ ≦

(D₂ /2), and

D₂ ≦

0.1.

11. The device according to claim 10, wherein:

D₁ ≦

0.2%; and

,

D₂ ≦

1.0%.

4. The method according to claim 4, wherein:

D₁ ≦

0.2%; and

,

D₂ ≦

1.0%.

6. The method according to claim 6, wherein said exposure time is selected to move either X₂ or X₁ to an extreme image density level.

12. A device for detecting and processing a specular image density level X₁ and a highlight image density level X₂ in a reproduction of an illuminated scene comprising:

means for exposing a photo-sensitive means by a luminous image from said scene, thereby generating a plurality of digital image signals, each said digital image signal representing a density of a portion of said luminous image;

means for establishing a cumulative histogram for the image density levels of said digital image signals;

means for providing a reference specular cumulative frequency Y₁ ;

means for providing a reference highlight cumulative frequency Y₂ ;

means for converting said specular cumulative frequency Y₁ via said cumulative histogram, to obtain said specular image density level X₁ ;

means for converting said highlight cumulative frequency Y₂ via said cumulative histogram, to obtain said highlight image density level X₂ ; and

means for indicating a presence of specular reflections when a ratio α

=∥

X_MAX -X₂ ∥

/∥

X_MAX -X₁ ∥

is smaller than a predetermined value, wherein X_MAX is an image density level greater than or equal to a highest density level within said luminous image.

13. The device according to claim 12, wherein said predetermined value is 0.875.

14. The device according to claim 12, further comprising means for optimizing an exposure time for exposing said photo-sensitive means based on the image density level of X₁, X₂ or a combination thereof.

15. The device according to claim 14, wherein said exposure time is selected to move either X₂ or X₁ to an extreme image level.

16. The device according to claim 12, wherein said image signals are further transformed by a gradation correction curve.

17. The device according to claim 16, wherein said gradation correction curve comprises linear segments, having a start point at X₁, a break point at X₂ and an end point at X_MAX.

18. The method according to claim 17, wherein said gradation correction curve comprises the break point is smoothed by an arc.