Information processing apparatus and method for calculating the position and orientation of an image sensing device

1. An information processing apparatus connected to an image sensing device which senses a movie of a physical space, comprising:

a calculation unit configured to calculate a plurality of pieces of candidate position and orientation information based on a plurality of pieces of sensing position and orientation information stored in a memory as candidates of a sensing position and a sensing orientation of the image sensing device upon sensing an image of each frame for respective frames which form the movie;

an evaluation unit configured to calculate evaluation values for the plurality of pieces of candidate position and orientation information;

a selection unit configured to select one candidate position and orientation information based on the evaluation values from the plurality of pieces of candidate position and orientation information; and

a recording unit configured to record the candidate position and orientation information selected by said selection unit in the memory as the sensing position and orientation information,

wherein said calculation unit calculates the plurality of pieces of candidate position and orientation information based on a plurality of motion models, and

wherein said evaluation unit comprises a correction unit configured to correct each candidate position and orientation information calculated by said calculation unit, and said evaluation unit calculates an evaluation value for the candidate position and orientation information corrected by said correction unit.

2. The apparatus according to claim 1 , wherein said calculation unit independently calculates a candidate position and a candidate orientation as candidate position and orientation information using identical or different motion models for a position and an orientation.

3. The apparatus according to claim 1 , wherein motions of the motion models include a static motion, uniform velocity motion, uniform acceleration motion, and random motion.

4. The apparatus according to claim 1 , wherein based on errors between detection positions of image features in an image of an acquired frame, and estimated positions of the image features on the image, which are estimated from candidate position information and candidate orientation information that configure the candidate position and orientation information and known three-dimensional information corresponding to the image features, said correction unit calculates correction values of the candidate position information and the candidate orientation information for each candidate position and orientation information, and corrects the candidate position and orientation information based on the calculated correction values.

5. The apparatus according to claim 4 , wherein said correction unit calculates weights according to magnitudes of the errors for respective image features and calculates the correction values based on the weights for each candidate position and orientation information.

6. The apparatus according to claim 5 , wherein said selection unit calculates, as the estimation value, an average value of the weights for each candidate position and orientation information, and selects candidate position and orientation information with a largest average value.

7. The apparatus according to claim 4 , wherein the image features are edges in the image, and the three-dimensional information is expressed by a set of edges.

8. The apparatus according to claim 1 , wherein said selection unit selects pieces of candidate position and orientation information, as many as a pre-set number, of the plurality of pieces of candidate position and orientation information, and

said recording unit records the pieces of candidate position and orientation information, as many as the pre-set number, which are selected by said selection unit, in the memory as the sensing position and orientation information.

9. The apparatus according to claim 8 , wherein said selection unit combines candidate position and orientation information, which is determined to have a position and an orientation close to a position and an orientation indicated by reference candidate position and orientation information, to the reference candidate position and orientation information as candidate position and orientation information to be combined, and

said selection unit selects pieces of candidate position and orientation information, as many as the pre-set number of pieces of candidate position and orientation information which have larger magnitudes of the evaluation values, of the remaining pieces of candidate position and orientation information other than the candidate position and orientation information to be combined.

10. An information processing method to be executed by an information processing apparatus, which is connected to an image sensing device which senses a movie of a physical space, comprising:

a calculation step of calculating a plurality of pieces of candidate position and orientation information based on a plurality of pieces of sensing position and orientation information stored in a memory as candidates of a sensing position and a sensing orientation of the image sensing device upon sensing an image of each frame for respective frames which form the movie;

an evaluation step of calculating evaluation values for the plurality of pieces of candidate position and orientation information;

a selection step of selecting one candidate position and orientation information based on the evaluation values from the plurality of pieces of candidate position and orientation information; and

a recording step of recording the candidate position and orientation information selected in the selection step in the memory as the sensing position and orientation information,

wherein said calculation step calculates the plurality of pieces of candidate position and orientation information based on a plurality of motion models, and

wherein said evaluation step comprises a correction step of correcting each candidate position and orientation information calculated in said calculation step, and said evaluation step calculates an evaluation value for the candidate position and orientation information corrected in said correction step.

11. A non-transitory computer-readable storage medium storing a computer for making a computer execute an information processing method according to claim 10 .