Apparatus and method for automatically analyzing digital video quality

• US 7,663,636 B2
• Filed: 05/22/2006
• Issued: 02/16/2010
• Est. Priority Date: 11/25/2005
• Status: Active Grant

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: Kim, Jin Seo, Cho, Maeng Sub, Koo, Bon Ki
• Primary Examiner(s): Wu; Xiao M
• Assistant Examiner(s): Merouan; Abderrahim
• Application Number:
  US11/437,611
• Publication Number:
  US 20070120863A1
• Time in Patent Office: 1,366 Days
• Field of Search: 345/581, 345582-618, 382/112, 382/199, 382/206, 382/228, 382/254, 382/266, 382/269, 382/272, 348 1401-843
• US Class Current: 345/581
• CPC Class Codes: H04N 17/004 for digital television systems

Abstract

An apparatus and method for automatically analyzing digital video quality are provided. The apparatus for automatically analyzing digital video quality includes a video analyzing unit and a result displaying unit. The video analyzing unit analyzes a texture attribute of an input video and analyzes a color attribute of the input video, thereby obtaining a texture attribute analysis result and a color attribute analysis result. The result displaying unit determines whether a reference video for video quality analysis exists or not. When the reference video exists, the result displaying unit determines the texture attribute analysis result as a main component analysis result and the color attribute analysis result as a minor component analysis result, respectively, and displays the determined analysis results on a screen in a form of numerical values and graphs together with the input video. On the contrary, when the reference video does not exist, the result displaying unit determines the texture attribute analysis result as the minor component analysis result and the color attribute analysis result as the major component analysis result, respectively, and displays the determined analysis results on the screen in a form of numerical values and graphs together with the input video. Accordingly, the digital video quality can be automatically analyzed and displayed.

Claims

1. An apparatus for automatically analyzing digital video quality, comprising:

a video analyzing unit configured to analyze a texture attribute of an input video and to analyze a color attribute of the input video to obtain a texture attribute analysis result and a color attribute analysis result, respectively, and assign the texture attribute analysis result and the color attribute analysis result to a main component analysis result and a minor analysis result based, upon the existence of a reference video, the textural attribute analysis result and the color attribute analysis result ranging from about zero, the lowest quality, to about 100, the highest quality; and

a result displaying unit configured to determine the existence of the reference video;

wherein when the reference video exists, the result displaying unit is configured;

to determine the texture attribute analysis result as a main component analysis result and the color attribute analysis result as a minor component analysis result, and

to display the determined analysis results on a screen in a form of numerical values and graphs together with the input video;

wherein, when the reference video does not exist, the results displaying unit is configured to determine the texture attribute analysis result as the minor component analysis result and the color attribute analysis result as the main component analysis result, respectively, and to display the determined analysis results on the screen in a form of numerical values and graphs together with the input video; and

wherein a final video quality is determined based upon the following equation;

Final Video Quality ={kΣ

(main component analysis result)+(1−

k)Σ

(minor component analysis result)}/3;

where k is a predetermined importance factor.

2. The apparatus of claim 1, wherein the texture attribute includes a surface reflection attribute, a surface texture attribute, and an edge attribute, and the color attribute includes lightness, chroma, and contrast.

3. The apparatus of the claim 1, wherein the video analyzing unit includes:

a texture analyzing unit for analyzing the texture attribute of objects existing in the input video by using a previously stored database and a frequency domain analysis; and

a color analyzing unit for analyzing the color attribute of R, G and B values for each pixel of the input video.

4. The apparatus of claim 3, wherein the texture analyzing unit includes:

a preprocessing unit configured to perform a basic analysis on the texture attribute of the input video by using the previously stored database and a video processing in the frequency domain;

a reflection analyzing unit configured to receive the basic analysis result from the preprocessing unit, analyze a surface reflection attribute of the objects existing in the input video, and output the analysis result in a form of normalized numerical values;

a surface texture analyzing unit configured to analyze the surface texture of the objects existing in the input video, and output the analysis result in a form of normalized numerical values; and

an edge analyzing unit configured to analyze the edge attribute of the objects existing in the input video, and output the analysis result in a form of normalized numerical values.

5. The apparatus of claim 3, wherein the color analyzing unit includes:

a preprocessing unit configured to convert the color values of the inputted signal into device-independent XYZ color space, and converting the color values of the XYZ color space into color values of standard color space;

a lightness analyzing unit configured to analyze lightness of the input video by using lightness values converted by the preprocessing unit;

a chroma analyzing unit configured to analyze chroma of the input video by using chroma values converted by the preprocessing unit; and

a contrast analyzing unit configured to analyze contrast of the input video by using the converted lightness values and the R, G and B color values of the input video.

6. The apparatus of claim 1, wherein the result displaying unit includes:

a video displaying unit configured to display the input video on a video quality analysis display screen; and

an analysis result displaying unit configured to display the main component analysis result, the minor component analysis result, and the final quality analysis result in a form of numerical values and graphs.

7. A method for automatically analyzing and displaying digital video quality, comprising the steps of:

(a) receiving a target video to be analyzed, and determining an importance factor k for determining weights of main and minor components to determine a final video quality;

(b) performing a preprocessing of a texture analysis, the preprocessing operation including calculating histogram, noise and reflection attribute of an input video and sequentially analyzing a texture attribute, a reflection attribute, and an edge attribute of the input video;

(c) converting R, G and B color values of the input video into device-independent XYZ color space in combination with the operation (b), converting the XYZ color space into color values of standard color domain, and sequentially analyzing a lightness attribute, a chroma attribute, and a contrast attribute of the input video;

(d) determining, based upon a textual attribute, a color attribute, and whether a reference video exists, a main component analysis result and a minor component analysis result, and calculating a final video quality analysis result by using the determined main and minor component analysis results and the importance factor k based upon the following equation;

Final Video Quality Analysis Result ={kΣ

(main component analysis result)+(1−

k)Σ

(minor component analysis result)}/3

wherein when the reference video exists, the main component analysis result and the minor component analysis result are based on the textual attribute and color attribute, respectively, and when the reference video does not exist, the main component analysis result and the minor component analysis result are based on the color attribute and textual attribute, respectively; and

(e) displaying the input video and the final video quality analysis result of the operation (d) in a form of numerical values and graphs.