Patent attributes
The present disclosure provides techniques for dimensioning an object by capturing two three-dimensional images of the object from orthogonal perspectives. For instance, a first three-dimensional image of a first side of a target object, and a second three-dimensional image of an orthogonal side of the generally cuboid object may be captured using a three-dimensional camera. A first two-dimensional axis-aligned minimum bounding box may be generated based on the first three-dimensional image and a second two-dimensional axis-aligned minimum bounding box may be generated based on the second three-dimensional image. The dimensions of the object may be calculated using the width of the first axis-aligned two-dimensional minimum bounding box, the width of the second axis-aligned two-dimensional minimum bounding box, and the greater of the height of the first axis-aligned two-dimensional minimum bounding box and the height of the second axis-aligned two-dimensional minimum bounding box.
