Image processing system, image processing method, and non-transitory computer readable medium for generating a colorized image

An image processing method comprising:

• acquiring a monochrome image including a target part;
• acquiring a category of the target part and a first color hint indicating a color of the target part;
• converting the first color hint into a second color hint indicating the color of the target part based on the acquired category;
• generating a colorized image corresponding to the monochrome image from the monochrome image and the second color hint of the target part by using a prediction model trained by machine learning;
• adjusting color space coordinates of the converted second color hint based on a distance between the color space coordinates of the converted second color hint and color space coordinates of the first color hint;
• and
• generating the colorized image corresponding to the monochrome image based on the adjusted second color hint.

A non-transitory computer readable medium storing an image processing program for causing a computer to:

• acquire a monochrome image including a target part;
• acquire a category of the target part and a first color hint indicating a color of the target part;
• convert the first color hint into a second color hint indicating the color of the target part based on the acquired category;
• generate a colorized image corresponding to the monochrome image from the monochrome image and the second color hint of the target part by using a prediction model trained by machine learning;
• adjust color space coordinates of the converted second color hint based on a distance between the color space coordinates of the converted second color hint and color space coordinates of the first color hint;
• and
• generate the colorized image corresponding to the monochrome image based on the adjusted second color hint.