Image processing method, image processing apparatus, image data compression apparatus, and computer program readable storage medium

An image processing method for expressing small regions in an image with not more than a predetermined number of representative colors, in which an image processing is carried out for each of the small regions, the method comprising:

• a first selection step of, with respect to a plurality of pixels in the small region, based on variation values that indicate the extent of variation, in the small region, of values of a plurality of color variables that express a color of each of the pixels, selecting one of the color variables as a first target color variable;
• a first reference value calculation step of, from among values from a maximum value to a minimum value of the first target color variable in the small region, when there is a value of the first target color variable for which the number of pixels is “0”, obtaining the value of the first target color variable as a first reference value for grouping the pixels;
• a first grouping step of grouping the pixels in the small region based on the first reference value obtained in the first reference value calculation step;
• a second selection step of determining the variation value for each color variable in each group obtained in the first grouping step, selecting the group with the largest variation value as a target group and selecting the color variable with the largest variation value as a second target color variable, based on the variation values of the color variables in each of the groups obtained in the first grouping step;
• a second reference value calculation step of, from among values from a maximum value to a minimum value of the second target color variable in the target group, when there is a value of the second target color variable for which the number of pixels is “0”, obtaining the value of the second target color variable in the target group as a second reference value for grouping the pixels;
• a second grouping step of grouping the pixels in the target group based on the second reference value obtained in the second reference value calculation step;
• and
• a representative color calculation step of, for each of groups obtained in the second grouping step, obtaining a representative color based on a value of each of the color variables of pixels that belong to the group.

An image processing method for expressing small regions in an image with not more than a predetermined number of representative colors, in which an image processing is carried out for each of the small regions, the method comprising:

• a first selection step of, with respect to a plurality of pixels in the small region, based on variation values that indicate the extent of variation, in the small region, of values of a plurality of color variables that express a color of each of the pixels, selecting one of the color variables as a target color variable;
• a first reference value calculation step of, from among values from a maximum value to a minimum value of the target color variable in the small region, when there is a value of the target color variable for which the number of pixels is “0”, obtaining the value of the target color variable as a first reference value for grouping the pixels, or in the alternative, when there is no value of the target color variable for which the number of pixels is “0”, obtaining a value nearest to a middle value between a maximum value and a minimum value as a first reference value;
• a first grouping step of grouping the pixels in the small region based on the first reference value obtained in the first reference value calculation step;
• and
• a first representative color calculation step of, for each of groups in the first-grouped small region, obtaining a representative color based on a value of each of the color variables of pixels that belong to the group
• wherein said plurality of color variables include a luminosity that has been corrected to increase the respective value of the luminosity when the originally obtained value for a specific pixel exceeds a set threshold.

An image processing apparatus that performs image processing that expresses small regions in an image with not more than a predetermined number of representative colors, the image processing being carried out for each of the small regions, the apparatus comprising:

• a first selection portion that, with respect to a plurality of pixels in the small region, based on variation values that indicate the extent of variation, in the small region, of values of a plurality of color variables that express a color of each of the pixels, selects one of the color variables as a first target color variable;
• a first reference value calculation portion that, from among values from a maximum value to a minimum value of the first target color variable in the small region, obtains a value of the first target color variable for which the number of pixels is “0”, as a first reference value for grouping the pixels;
• a first grouping portion that groups the pixels in the small region based on the first reference value obtained by the first reference value calculation portion;
• a second selection portion that, based on the variation values of the color variables in each of the groups obtained by the first grouping portion, determines the variation value for each color variable in each group obtained by the first grouping portion, selects the group with the largest variation value as a target group and selects the color variable with the largest variation value as a second target color variable;
• a second reference value calculation portion that, from among values from a maximum value to a minimum value of the second target color variable in the target group, when there is a value of the second target color variable for which the number of pixels is “0”, obtains the value of the second target color variable in the target group as a second reference value for grouping the pixels;
• a second grouping portion that groups the pixels in the target group based on the second reference value obtained by the second reference value calculation portion;
• and
• a representative color calculation portion that, for each of groups obtained by the second grouping portion, obtains a representative color based on a value of each of the color variables of pixels that belong to the group.

An image data compression apparatus, comprising:

• a partitioning portion that performs partitioning of image data for each of a plurality of small regions including a plurality of pixels;
• a first selection portion that, with respect to a plurality of pixels in the small region, based on variation values that indicate the extent of variation, in the small region, of values of a plurality of color variables that express a color of each of the pixels, selects one of the color variables as a first target color variable;
• a first reference value calculation portion that, from among values from a maximum value to a minimum value of the first target color variable in the small region, obtains a value of the first target color variable for which the number of pixels is “0” as a first reference value for grouping the pixels;
• a first grouping portion that groups the pixels in the small region based on the first reference value obtained by the first reference value calculation portion;
• a second selection portion that, based on the variation values of the color variables in each of the groups obtained by the first grouping portion, determines the variation value for each color variable in each group obtained by the first grouping portion, selects the group with the largest variation value as a target group and selects the color variable with the largest variation value as a second target color variable;
• a second reference value calculation portion that, from among values from a maximum value to a minimum value of the second target color variable in the target group, when there is a value of the second target color variable for which the number of pixels is “0”, obtains the value of the second target color variable in the target group as a second reference value for grouping the pixels;
• a second grouping portion that groups the pixels in the target group based on the second reference value obtained by the second reference value calculation portion;
• a representative color calculation portion that, for each of groups obtained by the second grouping, obtains a representative color based on a value of each of the color variables of pixels that belong to the group;
• and
• a compressed data generation portion that, with respect to each of the small regions of the image data, generates compressed data using the representative color obtained by the representative color calculation portion.

A non-transitory computer readable storage medium having stored thereon a computer program that causes image processing to be performed that expresses small regions in an image with not more than a predetermined number of representative colors, the image processing being carried out for each of the small regions, by causing a computer to execute:

• a first selection step of, with respect to a plurality of pixels in the small region, based on variation values that indicate the extent of variation, in the small region, of values of a plurality of color variables that express a color of each of the pixels, selecting one of the color variables as a first target color variable;
• a first reference value calculation step of, from among values from a maximum value to a minimum value of the first target color variable in the small region, obtaining a value of the first target color variable for which the number of pixels is “0” as a first reference value for grouping the pixels;
• a first grouping step of grouping the pixels in the small region based on first the reference value obtained in the reference value calculation step;
• a second selection step of determining the variation value for each color variable in each group obtained in the first grouping step, selecting the group with the largest variation value as a target group and selecting the color variable with the largest variation value as a second target color variable, based on the variation values of the color variables in each of the groups obtained by the first grouping portion;
• a second reference value calculation step of, from among values from a maximum value of the second target color variable in the target group, when there is a value of the second target color variable for which the number of pixels is “0”, obtaining the value of the second target color variable in the target group as a second reference value for grouping the pixels;
• a second grouping step of grouping the pixels in the target group based on the second reference value obtained by the second reference value calculation portion;
• and
• a representative color calculation step of, for each of groups obtained in the second grouping step, obtaining a representative color based on a value of each of the color variables of pixels that belong to the group.