According to one embodiment, an article recognition apparatus includes a first interface, a second interface, and a processor. The first interface acquires an image captured by photographing a photography range including an area where articles that are recognition targets are arranged. The second interface acquires depth information from a predetermined position relative to an area corresponding to the photography range. The processor extracts image areas of all products existing in the image by using the depth information, and determines whether an unidentified area due to overlapping of a plurality of the products exists in the image, based on the image areas of the products extracted from the image.