Device Apparatus and Method for 3d Image Interpolation Based on a Degree of Similarity Between a Motion Vector and a Range Motion Vector

1. A three-dimensional (3D) image interpolation device that performs frame interpolation on 3D video, the 3D image interpolation device comprising: a range image interpolation unit configured to generate at least one interpolation range image to be interpolated between a first range image and a second range image, the first range image indicating a depth of a first image included in the 3D video, and the second range image indicating a depth of a second image included in the 3D video; an image interpolation unit configured to generate at least one interpolation image to be interpolated between the first image and the second image; a range motion vector calculation unit configured to calculate, as a range motion vector, a motion vector between the first range image and the second range image; an image motion vector calculation unit configured to calculate, as an image motion vector, a motion vector between the first image and the second image; a vector similarity calculation unit configured to calculate a vector similarity that is a value indicating a degree of a similarity between the image motion vector and the range motion vector; and an interpolation parallax image generation unit configured to generate, based on the at least one interpolation image interpolated according to the vector similarity, at least one pair of interpolation parallax images having parallax according to a depth indicated by the at least one interpolation range image; and an interpolation image number determination unit configured to determine an upper limit of the number of interpolations, so that the number of the interpolations increases as the vector similarity calculated by the vector similarity calculation unit increases, wherein the interpolation parallax image generation unit is configured to generate the at least one pair of interpolation parallax images which is equal to or less than the upper limit determined by the interpolation image number determination unit.

2. The 3D image interpolation device according to claim 1 , wherein the range motion vector calculation unit is configured to calculate the range motion vector for each block having a first size, the image motion vector calculation unit is configured to calculate the image motion vector for each block having the first size, and the vector similarity calculation unit is configured to: (i) generate at least one of a histogram of directions of range motion vectors including the range motion vector and a histogram of powers of the range motion vectors, for each block having a second size greater than the first size; (ii) generate at least one of a histogram of directions of image motion vectors including the image motion vector and a histogram of powers of the image motion vectors, for each block having the second size; and (iii) calculate the vector similarity based on at least one of (a) a similarity between the histogram of the directions of the range motion vectors and the histogram of the directions of the image motion vectors and (b) a similarity between the histogram of the powers of the range motion vectors and the histogram of the powers of the image motion vectors.

3. The 3D image interpolation device according to claim 1 , wherein the interpolation image number determination unit is configured to determine, as the number of the interpolations, a number which is inputted by a user and is equal to or less than the upper limit, and the interpolation parallax image generation unit is configured to generate the at least one pair of interpolation parallax images which is equal to the number of the interpolations determined by the interpolation image number determination unit.

4. The 3D image interpolation device according to claim 1 , further comprising a range image obtainment unit configured to: (i) obtain the first range image based on a blur correlation between a plurality of captured images which are included in a first captured image group and have respective different focal distances; and (ii) obtain the second range image based on a blur correlation between a plurality of captured images which are included in a second captured image group and have respective different focal distances, the second captured image group being temporally subsequent to the first captured image group.

5. The 3D image interpolation device according to claim 4 , further comprising a texture image obtainment unit configured to: (i) obtain, as the first image, a first texture image by reconstructing one captured image included in the first captured image group based on blur information indicating a feature of blur in the one captured image; and (ii) obtain, as the second image, a second texture image by reconstructing one captured image included in the second captured image group based on blur information indicating a feature of blur in the one captured image.

6. The 3D image interpolation device according to claim 1 , wherein the 3D image interpolation device is implemented as an integrated circuit.

7. A 3D imaging apparatus, comprising: an imaging unit; and the 3D image interpolation device according to claim 1 .

8. A three-dimensional (3D) image interpolation method of performing frame interpolation on 3D video, the 3D image interpolation method comprising: generating at least one interpolation range image to be interpolated between a first range image and a second range image, the first range image indicating a depth of a first image included in the 3D video, and the second range image indicating a depth of a second image included in the 3D video; generating at least one interpolation image to be interpolated between the first image and the second image; calculating, as a range motion vector, a motion vector between the first range image and the second range image; calculating, as an image motion vector, a motion vector between the first image and the second image; calculating a vector similarity that is a value indicating a degree of a similarity between the image motion vector and the range motion vector; and generating, based on the at least one interpolation image interpolated according to the vector similarity, at least one pair of interpolation parallax images having parallax according to a depth indicated by the at least one interpolation range image; and determining an upper limit of the number of interpolations, so that the number of the interpolations increases as the vector similarity calculated by the vector similarity calculation unit increases, wherein the at least one pair of interpolation parallax images generated is equal to or less than the upper limit determined.

9. A non-transitory computer-readable recording medium for use in a computer, the recording medium having a computer program recorded thereon for causing the comptuer to execute the 3D image interpolation method according to claim 8 .