Image Processing Device, Image Processing System, Image Processing Method, and Image Processing Program

An image processing device comprising:

• a division circuit configured to divide an input image and output a plurality of first images;
• a first processor configured to:
• perform computation of an object detection model prepared in advance by providing each of the plurality of first images as inputs to the object detection model;
• and
• acquire a set of pieces of attribute information including an attribute value of an object included in each of the plurality of first images and a first region surrounding the object, as first metadata of the first images, the first region including coordinate information in the input image;
• a scaling circuit configured to output a second image obtained by shrinking the input image;
• a second processor configured to:
• perform computation of the object detection model by providing the second image as an input to the object detection model;
• and
• acquire a set of pieces of attribute information including the attribute value of the object included in the second image and a second region surrounding the object, as second metadata of the second image, the second region including the coordinate information in the input image;
• and
• a compositing processor configured to generate third metadata of the input image by combining the set of pieces of attribute information of the second metadata and a set of pieces of attribute information that are not held in common by the second metadata and the first metadata.

An image processing method comprising:

• dividing an input image and outputting a plurality of first images;
• performing computation of an object detection model prepared in advance by providing each of the plurality of first images as an input to the object detection model and acquiring a set of pieces of attribute information including an attribute value of an object included in each of the plurality of first images and a first region surrounding the object as first metadata of the first image, the first region including coordinate information in the input image;
• outputting a second image obtained by shrinking the input image;
• performing computation of the object detection model by providing the second image as an input to the object detection model and acquiring a set of pieces of attribute information including an attribute value of the object included in the second image and a second region surrounding the object as second metadata of the second image, the second region including the coordinate information in the input image;
• and
• generating third metadata of the input image by combining the set of pieces of attribute information of the second metadata and a set of pieces of attribute information that are not held in common by the second metadata and the first metadata.

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

• dividing an input image and outputting a plurality of first images;
• performing computation of an object detection model prepared in advance by providing each of the plurality of first images as an input to the object detection model and acquiring a set of pieces of attribute information including an attribute value of an object included in each of the plurality of first images and a first region surrounding the object as first metadata of the first image, the first region including coordinate information in the input image;
• outputting a second image obtained by shrinking the input image;
• performing computation of the object detection model by providing the second image as an input to the object detection model and acquiring a set of pieces of attribute information including an attribute value of the object included in the second image and a second region surrounding the object as second metadata of the second image, the second region including the coordinate information in the input image;
• and
• generating third metadata of the input image by combining the set of pieces of attribute information of the second metadata and a set of pieces of attribute information that are not held in common by the second metadata and the first metadata.