Image processing device, method of processing image, and image display apparatus

摘要

According to an embodiment, an image processing device includes: a first acquiring unit, a second acquiring unit, a first setting unit, a second setting unit, a first calculating unit, and a second calculating unit. The first acquiring unit acquires a plurality of captured images by imaging a target object from a plurality of positions. The second acquiring unit acquires a provisional three-dimensional position and a provisional size. The first setting unit sets at least one search candidate point near the provisional three-dimensional position. The second setting unit sets a search window for each projection position where the search candidate point is projected, the search window having a size. The first calculating unit calculates an evaluation value that represents whether or not the target object is included inside the search window. The second calculating unit calculates a three-dimensional position of the target object based on the evaluation value.

主权项

1. An image processing device comprising:
a first acquiring unit configured to acquire a plurality of captured images by imaging a target object from a plurality of positions; a second acquiring unit configured to acquire a provisional three-dimensional position of the target object from the plurality of the captured images and acquire a provisional size of the target object included in the plurality of the captured images, wherein the second acquiring unit
uses search windows having a plurality of search window sizes that are different from one another,scans each captured image, andacquires the provisional size of the target object included in a scanned captured image; a first setting unit configured to set at least one search candidate point in a vicinity of the provisional three-dimensional position in a real space; a second setting unit configured to set search windows used to search the target object, the search windows corresponding to each projection position at which the search candidate point is projected on each captured image, the search windows having a size that is based on the position of the search candidate point and the provisional size; a first calculating unit configured to calculate evaluation values representing whether or not the target object is included inside the search windows; and a second calculating unit configured to calculate a three-dimensional position of the target object by using at least one search candidate point based on the evaluation values.