HMMD color space and method for quantizing color using HMMD space and color spreading
AssigneeLG Electronics Inc.
InventorsHyeon Jun Kim, Jin Soo Lee
Life of this patent
Estimating the term…
Abstract
A new color model HMMD is disclosed based upon hue, the shade, the tone, the tint and the brightness of a color, and a color quantizing method using the HMMD color space. The present invention regards the colors adjacent to a given color in the color space as the neighboring colors to consider indices of the neighboring colors as the partial values according to the distance. Accordingly, the present invention can accomplish a color quantization close to the change of the color sensed by the human eye, thereby capable of enhancing a performance of the image searching system based on content.
Claims15
A HMMD color model having a shape which includes two symmetrical cones merged at flat surfaces, said merged cones including an axis disposed between an apex A and an apex B, comprising:
- a hue defined as an angle around the AB axis;
- a diff defined as a vector normal to the AB axis in a direction from the AB axis to a circumference C of the cones;
- a max value defined as a vector in a direction from the apex B of a bottom one of the cones to the circumference C;
- a min value defined as a vector in a direction from the circumference C to the apex A of a top one of the cones;
- and
- a sum value defined as a vector in a positive direction of the AB axis,
- wherein each of the max value, the min value, the diff, and the sum value range between 0 and 1 and the hue ranges from 0 to 360, and
- wherein the HMMD color model is converted from an RGB color model with (r, g, b) color values in accordance with the following equations:
- max value=maximum value from input r, g, b values;
- min value=minimum value from the input r, g, b values;
- diff=max value−min value;
- sum=(max+min)/2;
- and
- hue=
- (g−b)/(max−min)*60, if (r=max∩(g−b)≧0)
- (g−b)/(max−min)*60+360, if (r=max∩(g−b)<0)
- (2.0+(b−r))/(max−min)*60, if (g=max)
- (4.0+(r−g))/(max−min)*60, if (g=max)
- 0, if max=min.
A quantization method for an HMMD color space, comprising:
- (a) dividing a gray area of the HMMD color space;
- (b) color quantizing an MMD space;
- (c) color quantizing a hue area;
- and
- (d) mapping color information within each of the quantized areas to an identical value, wherein step (b) includes:
- dividing the MMD space by at least one line parallel to a center axis of the HMMD color space;
- dividing the MMD space by at least one line parallel to a max vector of the MMD space;
- and
- dividing the MMD space by at least one line parallel to a min vector of the MMD space, wherein the at least one line parallel to the center axis of the HMMD color space and the at least one line parallel to the max vector of the MMD space are reference lines separating different color groups, and wherein the dividing steps result in the MMD space being divided into a plurality of triangles.
A quantization method for an HMMD color space, comprising:
- (a) dividing a gray area of the HMMD color space;
- (b) color quantizing an MMD space;
- (c) color quantizing a hue area;
- and
- (d) mapping color information within each of the quantized areas to an identical value, wherein step (d) includes:
- determining concentric points of each of the quantized areas;
- determining a distance of a color position in a first quantized area from a concentric point of the first quantized area;
- determining a distance of the color position from a concentric point of a second adjacent quantized area;
- allocating weighted values to the first quantized area and to the second adjacent quantized area based upon the determined distances;
- and
- mapping the color position according to allocated weighted values,
- wherein in the step of allocating weighted values, the weights are allocated by the equation:
- W i = 1 - d i ∑ n = 0 m d n
- where Wi is the weight for the (i)th quantized area, di is the distance of the color position in the (i)th quantized area from the concentric point of the (n)th quantized area, dn to dm are the distances of the color position from the concentric points of the quantized area or areas adjacent the first quantized area.
A quantization method for a color space having a plurality of color groups, comprising:
- determining concentric points of each of said color groups;
- determining a distance of a color position in a first color group from the concentric point of the first color group;
- determining distances of the color position from the concentric points of the color group or groups adjacent the first color group;
- allocating weighted values to the first color group and to the color group or groups adjacent the first color group based on the determined distances;
- and
- mapping the color according to the allocated weighted values,
- wherein in the step of allocating weighted values, the weights are allocated by the equation:
- W i = 1 - d i ∑ n = 0 m d n
- where Wi is the weight for the (i)th color group, di is the distance of the color position in the (i)th color group from the concentric point of the (n)th quantized area, dn to dm are the distances of the color position from the concentric points of the color group or groups adjacent the first color group.
Dependencies are inferred from the claim text.