Image processing apparatus, image processing method, and storage medium
An image processing apparatus includes a detection unit configured to detect a specific area from a captured image, and a determination unit configured to determine, based on the detected specific area, a light source color including a characteristic color by using a learning unit trained in advance by machine learning.
1. An image processing apparatus comprising:
one or more processors configured to function as following units:
a detection unit configured to detect a specific area from a captured image;
a first determination unit configured to determine, based on a color included in the detected specific area, a light source color including a characteristic color by using a learning unit trained in advance by machine learning;
a first calculation unit configured to calculate a first white balance correction value corresponding to the light source color determined by the first determination unit;
a second determination unit configured to determine the light source color based on a color of an area including an area other than the specific area;
a second calculation unit configured to calculate a second white balance correction value corresponding to the light source color determined by the second determination unit; and
a third calculation unit configured to calculate a third white balance correction value based on the first white balance correction value and the second white balance correction value.
2. The image processing apparatus according to claim 1 , wherein the one or more processors are further configured to function as a correction unit configured to perform a white balance correction on the captured image based on the third white balance correction value.
3. The image processing apparatus according to claim 1 , wherein the second calculation unit calculates the second white balance correction value corresponding to the light source color determined based on a color of a white area.
4. The image processing apparatus according to claim 1 , wherein the third calculation unit generates the third white balance correction value by mixing the first white balance correction value and the second white balance correction value at a mixing ratio based on reliability of the determined light source color.
5. The image processing apparatus according to claim 1 , wherein the specific area is an area including the characteristic color, corresponding to a known object, and having a characteristic color distribution.
6. The image processing apparatus according to claim 5 , wherein the characteristic color is a human skin color.
7. The image processing apparatus according to claim 1 , wherein learning data for the learning unit is data having a correlation between a light source having a color distribution preliminarily distributed at a specific distance from a blackbody radiation locus and an object having the characteristic color.
8. The image processing apparatus according to claim 1 ,
wherein the one or more processors are further configured to function as a processing unit configured to perform conversion processing to reduce an effect of an individual difference of an image capturing apparatus configured to capture an image, and
wherein a signal value to be input to the learning unit is a signal value subjected to the conversion processing.
9. The image processing apparatus according to claim 8 , wherein the processing unit converts red (R), green (G), and blue (B) values for the specific area corresponding to coordinates of a space with the individual difference into R, G, and B values corresponding to coordinates of a space without the individual difference based on a distance between two straight lines obtained by performing linear interpolation between two light sources determined for each of the spaces based on the R, G, and B values for the specific area, from among a plurality of light sources of which the R, G, and B values corresponding to the space without the individual difference and the R, G, and B values corresponding to the space with the individual difference are known.
10. An image processing method comprising:
detecting a specific area from a captured image;
determining, in a first determination based on a color included in the detected specific area, a light source color including a characteristic color by using a learning unit trained in advance by machine learning;
calculating a first white balance correction value corresponding to the light source color determined in the first determination;
determining, in a second determination, the light source color based on a color of an area including an area other than the specific area;
calculating a second white balance correction value corresponding to the light source color determined in the second determination; and
calculating a third white balance correction value based on the first white balance correction value and the second white balance correction value.
11. A non-transitory computer-readable storage medium storing a program for causing a computer to execute an image processing method comprising:
detecting a specific area from a captured image;
determining, in a first determination based on a color included in the detected specific area, a light source color including a characteristic color by using a learning unit trained in advance by machine learning;
calculating a first white balance correction value corresponding to the light source color determined in the first determination;
determining, in a second determination, the light source color based on a color of an area including an area other than the specific area;
calculating a second white balance correction value corresponding to the light source color determined in the second determination; and
calculating a third white balance correction value based on the first white balance correction value and the second white balance correction value.