摘要 |
A commodity recognition apparatus extracts the appearance feature amount of a commodity included in an image from the captured image, compares the appearance feature amount with the feature amount data of a recognition dictionary, and extracts the candidate of the commodity included in the image. When the selection input of the commodity is accepted from a plurality of commodity candidates, correction data are generated according to the feature amount data of the commodity of which the selection input is accepted and the appearance feature amount data, and the appearance feature amount of the commodity extracted by utilizing the correction data is corrected. |