Image processing apparatus, method and program for determining arrangement of vectors on a distribution map

摘要

There is provided an image processing apparatus that includes a feature data calculating unit, a distribution map generating unit, and a display control unit. The feature data calculating unit calculates feature data of each image contained in a group of sequential images that are taken by an image taking apparatus. The distribution map generating unit includes a representative vector calculating unit and a representative vector arrangement determining unit, and generates a distribution map that represents an overall trend of the group of sequential images. The representative vector calculating unit calculates a predetermined number of representative vectors that represent a plurality of images contained in the group of sequential images. The representative vector arrangement determining unit determines arrangement of the calculated representative vectors on the distribution map. The display control unit displays the distribution map on a display unit.

主权项

1. An image processing apparatus, comprising:
a processor; and a storage unit storing a group of sequential images that are taken by an image taking apparatus; wherein the processor is configured to interact with the storage unit to calculate feature data of each image contained in the group of sequential images; wherein the processor is configured to interact with the storage unit to generate, by a learning process using the feature data of each image, a distribution map that represents an overall trend of the group of sequential images; wherein the processor is configured to interact with the storage unit to display the distribution map on a display unit; wherein the processor, when configured to generate the distribution map, is configured to calculate a predetermined number of representative vectors that represent a plurality of images contained in the group of sequential images, during the learning process using the feature data of each image, based on the feature data of each image and a level of importance of each image by giving more weight in the learning process to the feature data of an important image than the feature data of the other images when the important image has a level of importance higher than a predetermined value; and wherein the processor is configured to determine arrangement of the calculated representative vectors on the distribution map.