According to an embodiment, an image processing apparatus includes a first extracting unit, a second extracting unit, and a detecting unit. The first extracting unit is configured to extract a first region in which an object is represented, from a distance image that is drawn using distance information, based on the distance information of the object calculated from an image of the object captured by an image capturing unit. The second extracting unit is configured to extract a contour direction that is a direction along which pixels forming a contour of the first region are arrayed. The detecting unit is configured to detect a first surface facing the image capturing unit from the first region, based on the contour direction extracted by the second extracting unit.