CPC H04N 1/6069 (2013.01) [G06T 7/10 (2017.01); G06T 7/90 (2017.01); G06V 10/764 (2022.01); G06V 10/774 (2022.01); G06V 10/82 (2022.01); G06V 20/70 (2022.01); H04N 1/00336 (2013.01); H04N 1/54 (2013.01); H04N 1/6072 (2013.01); G06T 2207/10024 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] | 20 Claims |
1. A computer-implemented method for object color management by a raster image processor and an object detection and semantic segmentation module, the method comprising:
receiving, by a computer system, an image for depositing ink onto a substrate to print the image on the substrate,
wherein the image comprises at least one object;
performing semantic segmentation, by the object detection and semantic segmentation module, on the image to extract the at least one object from the image,
wherein the object detection and semantic segmentation module is trained to perform the semantic segmentation using multiple training images;
generating, by the object detection and semantic segmentation module, (1) at least one mask layer corresponding to the at least one object and (2) at least one spot color for the at least one object based on the multiple training images;
generating, by the raster image processor, (3) at least one process ink layer based on the image and (4) at least one spot color layer based on the at least one mask layer and the at least one spot color,
wherein the at least one process ink layer is associated with at least one color of the ink;
generating, by the raster image processor, print data based on the at least one process ink layer and the at least one spot color layer; and
sending, to a printing system, the print data for printing the image on the substrate.
|