KR980007733A - A method for determining the visibility of each video object and a high-efficiency image encoding device using the same - Google Patents

Abstract

The present invention relates to a method for determining the visibility of each object and a moving picture encoding apparatus using the same. In the present invention, each characteristic is classified by using a visual sensitivity characteristic such as a visual sensitivity effect, a visual color difference range, and the like on a pixel which is a minimum unit constituting a video object constituting a digital image, Determines the human visibility of the image objects. Thus, the human being thus determined reflects the visibility into the quantization step size determination, so that the quantization step size can be determined so as to minimize the stitching quantization error according to the sensitivity of each video object at the time of quantization, thereby minimizing deterioration in image quality of the restoration projection.

Description

A method for determining visibility of a video object and a high-performance video encoding apparatus using the same

Since this is a trivial issue, I did not include the contents of the text.

FIG. 1 is a block diagram showing a general moving image encoder. FIG.

FIG. 2 is a block diagram showing an apparatus for realizing a visual sensitivity determination method for each video object according to the present invention; FIG.

FIG. 3 is a view showing an example of constructing a visual color difference lookup table used in the present invention; FIG.

4A and 4B are diagrams for explaining the operation of the visual sense complexity determining unit in FIG. 2;

FIG. 5 is a block diagram of a high-performance moving picture encoding apparatus using a method for determining the visibility of each video object according to the present invention.

Claims (9)

YU color coordinate system to generate a colorimetric color difference lockup table representing a colorimetric color difference tolerance range based on a human's colorimetric color difference characteristic by inputting a digital image constituted by a YU color coordinate system, An object classification process for classifying image objects in the digital image using a time-and-view color-difference allowable range; a brightness / contrast determination process for determining a brightness sensitivity between the classified image objects using a brightness / contrast effect; Determining a color sensitivity for each of the image objects using an average allowable color difference of the image objects, and determining a color sensitivity of the image objects using the average quantization step size and the color sense lock- Color Difference Determination Process: Using the color difference relationship between neighboring pixels Determining a visibility complexity of the image object; and determining a human visual sensitivity of the image object using brightness sensitivity, color sensitivity, quantization color change degree, and visibility complexity of the image objects. And determining a visibility of each video object. 2. The method of claim 1, wherein in the image input step, the sensory color difference lock-up table divides YUV chromaticity coordinates into 8 equal parts in a YUV chromaticity coordinate system to divide the YUV chromaticity coordinates into 512 total representative colors: , U, and V color coordinates of the YUV chromaticity coordinates of each color component of the digital projection, and determining YUV chromaticity coordinates for each pixel of the digital projection, Wherein the step of determining the visual sensitivity of each video object comprises the steps of: The method as claimed in claim 2, wherein, in the image inputting step, the allowable range of the visible color difference for each of the Y, U, and V color components is converted into the CIE L * a * b * uniform perceived color space during the YUV unbalance; Defining a color difference tolerance range for L, a, and b components in the converted CIE L * a * b * color space; And converting the allowable color difference tolerance range for each of the L, a, b components into a visible color difference tolerance range for the Y, U, V color components. The method of claim 1, wherein the object classifying step comprises: defining a color difference tolerance range for each of the Y color components that can be expressed from 16 to 235 using a color value defined as U = 128 and V = 128; Classifying image objects according to a Y color component value of a digital image belonging to the allowable range of color difference and reconstructing the digital image for each image object. The method as claimed in claim 1, wherein the step of determining a brightness / contrast ratio comprises: determining an average brightness of the digital image; determining that the brightness of the image object is 1.2 times brighter than the average brightness; The degree of visibility of the image objects determined to be present is determined using the number of pixels having a Y value greater than 1.2 times the average brightness in the image object and the number of all pixels constituting the digital image The method comprising the steps of: The method of claim 1, wherein the step of determining a color difference comprises: obtaining a minimum allowable color difference, a maximum allowable color difference, and an average allowable color difference of the image object; Obtaining a color sensitivity of the image object; calculating a maximum quantization error that can be generated by the average quantization step size when quantizing the digital image; Calculating a number of pixels having a permissible color difference smaller than the maximum quantization error by calculating a number of pixels having a permissible color difference smaller than the maximum quantization error and a total number of pixels constituting the digital image; And determining the number of pixels based on the number of pixels. The method of claim 1, wherein the step of determining a visually perceived complexity comprises: determining a maximum number of visible and visible chrominance occurrences that can be defined in a valid region in a color object comparison in the image object; Calculating a first allowable visual color difference of a pixel to be compared at present; calculating a second allowed visual color difference for each of a maximum of four comparison pixels defined in a predetermined direction; Comparing the first and second allowable visual color differences to calculate a number of visual and color difference occurrences by counting a case where the first allowable visual color difference is smaller than the second allowable visual color difference; Determining a visual acuity complexity using the maximum number of visual appearance chrominance occurrences and the number of visual acustic chrominance occurrences; 2. The method of claim 1, wherein the visibility determination step comprises: adjusting a degree of visibility of the image object and a magnitude of a degree of color change upon quantization according to a predetermined experimental value; Wherein the number of all pixels in the image object is determined based on a relative size between image objects determined according to the number of all pixels constituting the digital image, Determining a degree of distribution of the at least one object; And determining a human visual sensitivity to the image object using the degree of distribution of pixels sensitive to the human sensation, the color sensitivity determined in the sensory color difference determination process, and the sensory complexity determined in the sensory complexity determination process And determining a visual sensitivity for each video object. A digital image to be encoded is input and classified into I. P. B pictures, and a current frame image and a previous frame reconstructed image are supplied to an image input unit; A prediction error generator for estimating a motion vector between the current frame image and the previous frame reconstructed image and generating a prediction error image using the estimated motion vector; A quantizer for compressing an actual image signal by performing quantization with a predetermined quantization step size; a bit generator for encoding the image signal compressed by the quantizer using a variable length code table; An object-by-object sensitivity determining unit for determining a human's visual sensitivity for each of the image objects from the sensory-color-difference characteristic value and the sensory-complexity characteristic value; A bit adjustment section for adjusting the quantization step size according to the temporal sensitivity, Wherein the high-efficiency moving picture coding apparatus comprises: