Main color extracting method

Abstract

A primary color extraction method includes inputting an image; converting the pixel count of all color units in the image into a color histogram; determining a plurality of representative colors from the color histogram based on the pixel count of the color units, wherein after each representative color is determined, the pixel count of the color histogram is attenuated based on the color difference with the determined representative color; and classifying the plurality of representative colors into a plurality of groups based on the color difference using a k-average algorithm, and determining a plurality of primary colors from the plurality of groups.

Claims

1. A primary color extraction method, comprising: inputting an image; converting the pixel count of all color units in the image into a color histogram; determining a plurality of representative colors from the color histogram based on the pixel count of the color units, wherein, After each representative color is determined, the number of pixels in the color histogram is reduced based on the color difference with the determined representative color; and a k-means algorithm is used to classify the multiple representative colors into multiple groups based on the color difference, and multiple primary colors are determined from the multiple groups.

2. The primary color extraction method as described in claim 1, wherein the k-average algorithm uses at least one of the plurality of representative colors as the initial center.

3. The primary color extraction method as described in claim 1, wherein the step of dividing the plurality of representative colors into the plurality of groups using the k-means algorithm comprises: performing a weighted processing on the plurality of representative colors, and then grouping the plurality of representative colors using the k-means algorithm.

4. The primary color extraction method as described in claim 3, wherein the weighting process is a weighted square of the number of pixels of the plurality of representative colors.

5. The primary color extraction method as described in claim 1, further comprising: performing an image erosion step on the image and the corresponding plurality of groups to obtain a candidate color group; performing an adjustment procedure on the candidate color group; and determining the primary color from the candidate color group.

6. The primary color extraction method as described in claim 1, wherein the attenuation process is the number of pixels of all color units in the color histogram, where is the color difference between a color unit and one of the determined representative colors, and is a constant.

Images