Monochromatic image determination
View Patent ↗Systems, methods and computer readable media for determination of monochromatic images are described. Some implementations can include a method. The method can include converting an image in a first colorspace to a first converted image in a second colorspace. The method can also include generating a hue histogram based on hue values in the converted image and determining a hue dispersion measure based on the hue histogram and a number of pixels in the image. The method can further include determining a hue clustering value based on the hue histogram and converting the image in the first colorspace into a second converted image in a third colorspace. The method can also include determining a variance measure based on the second converted image, and comparing the hue dispersion measure, the hue clustering value and the variance measure to a first threshold value, a second threshold value and a third threshold value, respectively.
1. A method comprising:
converting an image in a first colorspace to a first converted image in a second colorspace;
generating a hue histogram based on hue values in the converted image;
determining a hue dispersion measure based on the hue histogram and a number of pixels in the image;
determining a hue clustering value based on the hue histogram;
converting the image in the first colorspace into a second converted image in a third colorspace;
determining a variance measure based on the second converted image;
comparing the hue dispersion measure, the hue clustering value and the variance measure to a first threshold value, a second threshold value and a third threshold value, respectively; and
providing an indication of whether the image is monochromatic based on one or more of the comparisons,
wherein determining the hue dispersion measure includes:
projecting hue values onto a unit circle, wherein any pixel missing from the histogram is assigned to point (0,0);
computing a center of gravity for plotted pixels;
iterating over plotted pixels on the unit circle and computing an intermediate value according to (hx−x)^8+(hy−y)^8 for each plotted pixel;
summing the intermediate values; and
normalizing the sum of intermediate values, wherein the normalized result is the hue dispersion measure.
2. The method of claim 1 , wherein the second colorspace is an HSV colorspace.
3. The method of claim 1 , wherein the third colorspace is a YCbCr colorspace.
4. The method of claim 3 , wherein determining the variance measure includes:
building histograms of Cb and Cr values for each Y value;
summing corresponding variances per Y value; and
multiplying image Cb and Cr variances, wherein the result of multiplication is the CbCr variance measure.
5. The method of claim 1 , wherein determining the hue clustering value includes:
discarding hue histogram values within about 30 degrees from the largest hue histogram value;
storing the largest hue histogram value as h, and a degree corresponding to the largest hue histogram value as m;
when a next largest hue histogram value is larger or equal to alpha*h, processing the next largest histogram value that has not been discarded and increasing the hue clustering value of the image the degree distance between m and a degree d of the next largest hue histogram value on a circle; and
iterating the processing until a next largest hue histogram value is reached that is not greater than or equal to alpha*h.
6. The method of claim 1 , wherein the image is a video frame and the method further includes determining whether a plurality of images corresponding to a selected number of frames are monochromatic and merging results of the selected number of frames together to generate an indication of whether the video is monochromatic.
7. A system comprising one or more computers configured to perform operations including:
converting an image in a first colorspace to a first converted image in a second colorspace;
generating a hue histogram based on hue values in the converted image;
determining a hue dispersion measure based on the hue histogram and a number of pixels in the image;
determining a hue clustering value based on the hue histogram;
converting the image in the first colorspace into a second converted image in a third colorspace;
determining a variance measure based on the second converted image;
comparing the hue dispersion measure, the hue clustering value and the variance measure to a first threshold value, a second threshold value and a third threshold value, respectively; and
providing an indication of whether the image is monochromatic based on one or more of the comparisons,
wherein determining the hue clustering value includes:
discarding hue histogram values within about 30 degrees from the largest hue histogram value;
storing the largest hue histogram value as h, and a degree corresponding to the largest hue histogram value as m;
when a next largest hue histogram value is larger or equal to alpha*h, processing the next largest histogram value that has not been discarded and increasing the hue clustering value of the image the degree distance between m and a degree d of the next largest hue histogram value on a circle; and
iterating the processing until a next largest hue histogram value is reached that is not greater than or equal to alpha*h.
8. The system of claim 7 , wherein the second colorspace is an HSV colorspace.
9. The system of claim 7 , wherein the third colorspace is a YCbCr colorspace.
10. The system of claim 9 , wherein determining the variance measure includes:
building histograms of Cb and Cr values for each Y value;
summing corresponding variances per Y value; and
multiplying image Cb and Cr variances, wherein the result of multiplication is the CbCr variance measure.
11. The system of claim 7 , wherein determining the hue dispersion measure includes:
projecting hue values onto a unit circle, wherein any pixel missing from the histogram is assigned to point (0,0);
computing a center of gravity for each plotted point;
iterating over plotted pixels on the unit circle and computing an intermediate value according to (hx−x)^8+(hy−y)^8 for each plotted pixel;
summing the intermediate values; and
normalizing the sum of intermediate values, wherein the normalized result is the hue dispersion measure.
12. The system of claim 7 , wherein the image is a video frame and the method further includes determining whether a plurality of images corresponding to a selected number of frames are monochromatic and merging results of the selected number of frames together to generate an indication of whether the video is monochromatic.
13. A nontransitory computer readable medium having stored thereon software instructions that, when executed by a processor, cause the processor to perform operations including:
converting an image in a first colorspace to a first converted image in a second colorspace;
generating a hue histogram based on hue values in the converted image;
determining a hue dispersion measure based on the hue histogram and a number of pixels in the image;
determining a hue clustering value based on the hue histogram;
converting the image in the first colorspace into a second converted image in a third colorspace;
determining a variance measure based on the second converted image;
comparing the hue dispersion measure, the hue clustering value and the variance measure to a first threshold value, a second threshold value and a third threshold value, respectively; and
providing an indication of whether the image is monochromatic based on one or more of the comparisons,
wherein the third color space is a YCbCr color space, and
wherein determining the variance measure includes:
building histograms of Cb and Cr values for each Y value;
summing corresponding variances per Y value; and
multiplying image Cb and Cr variances, wherein the result of multiplication is the CbCr variance measure.
14. The nontransitory computer readable medium of claim 13 , wherein the second colorspace is an HSV colorspace.
15. The nontransitory computer readable medium of claim 13 , wherein determining the hue dispersion measure includes:
projecting hue values onto a unit circle, wherein any pixel missing from the histogram is assigned to point (0,0);
computing a center of gravity for each plotted point;
iterating over plotted pixels on the unit circle and computing an intermediate value according to (hx−x)^8+(hy−y)^8 for each plotted pixel;
summing the intermediate values; and
normalizing the sum of intermediate values, wherein the normalized result is the hue dispersion measure.
16. The nontransitory computer readable medium of claim 13 , wherein determining the hue clustering value includes:
discarding hue histogram values within about 30 degrees from the largest hue histogram value;
storing the largest hue histogram value as h, and a degree corresponding to the largest hue histogram value as m;
when a next largest hue histogram value is larger or equal to alpha*h, processing the next largest histogram value that has not been discarded and increasing the hue clustering value of the image the degree distance between m and a degree d of the next largest hue histogram value on a circle; and
iterating the processing until a next largest hue histogram value is reached that is not greater than or equal to alpha*h.