Image-processing device, image-processing method, and recording medium
In order to produce a discriminator that has higher discrimination ability, this image-processing device is provided with a synthesis unit for synthesizing a background image and an object image the hue and/or brightness of which at least partially resembles at least a portion of the background image, a generation unit for generating a difference image between the synthesized image and the background image, and a machine learning unit for performing machine learning using the generated difference image as learning data.
1. An image-processing device comprising:
at least one memory storing instructions; and
at least one processor configured to execute the instructions to:
generate a difference image between a background image and a synthesized image wherein the synthesized image is generated by synthesizing the background image and an object image having at least one portion close in at least one of hue, saturation and brightness to at least one portion of the background image;
perform machine learning using the difference image as learning data;
add noise to the synthesized image;
generate a difference image between the background image and the synthesized image to which the noise has been added; and
add a pseudo-shadow as noise to the synthesized image by transforming the object image, inferring a shadow part segment of an object in the synthesized image by using the synthesized image and the transformed object image, and altering a luminance of the shadow part segment of the synthesized image in relation to the synthesized image.
2. The image-processing device according to claim 1 , wherein the noise is at least one of a pseudo-shadow, impulse noise, and Gaussian noise.
3. The image-processing device according to claim 1 ,
wherein the at least one processor is further configured to apply a blur to the background image and the synthesized image, and
generate a difference image between the background image and the synthesized image to which the blur has been applied.
4. The image-processing device according to claim 1 ,
wherein the at least one processor is further configured to apply a blur to the generated difference image, and
perform machine learning using the difference image to which the blur has been applied as learning data.
5. The image-processing device according to claim 1 , wherein the object image is a human figure image.
6. The image-processing device according to claim 1 , wherein a neural network is used for the machine learning.
7. An image-processing method comprising:
generating a difference image between a background image and a synthesized image wherein the synthesized image is generated by synthesizing the background image and an object image having at least one portion close in at least one of hue, saturation and brightness to at least one portion of the background image;
performing machine learning using the difference image as learning data;
adding noise to the synthesized image;
generating a difference image between the background image and the synthesized image to which the noise has been added; and
adding a pseudo-shadow as noise to the synthesized image by transforming the object image, inferring a shadow part segment of an object in the synthesized image by using the synthesized image and the transformed object image, and altering a luminance of the shadow part segment of the synthesized image in relation to the synthesized image.
8. A non-transitory computer-readable recording medium storing a program configured to cause a computer to execute:
a process of generating a difference image between a background image and a synthesized image wherein the synthesized image is generated by synthesizing the background image and an object image having at least one portion close in at least one of hue, saturation and brightness to at least one portion of the background image;
a process of performing machine learning using the difference image as learning data;
a process of adding noise to the synthesized image;
a process of generating a difference image between the background image and the synthesized image to which the noise has been added; and
a process of adding a pseudo-shadow as noise to the synthesized image by transforming the object image, inferring a shadow part segment of an object in the synthesized image by using the synthesized image and the transformed object image, and altering a luminance of the shadow part segment of the synthesized image in relation to the synthesized image.