Arithmetic processing device

1. An arithmetic processing device comprising:

a processor that:

(i) acquires a first viewpoint signal and a second viewpoint signal relating to images composed of a plurality of pixels;

(ii) divides a first image represented by the acquired first viewpoint signal into a plurality of blocks each including a predetermined number of pixels;

(iii) sets a first predetermined number of first blocks at first intervals in a second image represented by the second viewpoint signal, calculates a first evaluated value indicating a degree of matching by matching a matching target block in the first image with each of the first blocks, selects one of the first blocks from the first predetermined number of the first blocks based on the first evaluated value, and calculates a first parallax between the selected first block and the matching target block;

(iv) sets a second predetermined number of second blocks at second intervals in a second image represented by the second viewpoint signal, calculates a second evaluated value indicating a degree of matching by matching the matching target block with each of the second blocks, selects one of the second blocks from the second predetermined number of the second blocks based on the second evaluated value, and calculates a second parallax between the selected second block and the matching target block; and

(v) determines, based on the first evaluated value and the second evaluated value and based on the first parallax and the second parallax, whether or not to employ one of the first parallax and the second parallax as a parallax between the matching target block and a block in the second image corresponding to the matching target block,

wherein the processor:

(i) multiplies each of the first evaluated values obtained by the calculation by a first weighting value which is different for each first evaluation value, and selects one of the first blocks as a matching candidate based on a result of the multiplication,

(ii) multiplies each of the second evaluated values obtained by the calculation by a second weighting value which is different for each second evaluated value, and selects one of the second blocks as the matching candidate based on a result of the multiplication,

wherein the first weighting values of the first blocks take a greater value as a position of the first block corresponding to the first weighting value is closer to a central position of a first range being defined by the first predetermined number and the first interval, and

the second weighting values of the second blocks take a smaller value as a position of the second block corresponding to the second weighting value is closer to a central position of a second range being defined by the second predetermined number and the second interval.

2. The arithmetic processing device according to claim 1 , wherein the processor:

(i) employs neither of the first parallax and the second parallax as the parallax between the matching target block and the block in the second image corresponding to the matching target block when both of the first evaluated value and the second evaluated value are smaller than a reference value,

(ii) employs the first parallax as the parallax between the matching target block and the block in the second image corresponding to the matching target block when the first evaluated value is greater than or equal to the reference value and the second evaluated value is smaller than the reference value,

(iii) employs the second parallax as the parallax between the matching target block and the block in the second image corresponding to the matching target block when the first evaluated value is smaller than the reference value and the second evaluated value is greater than or equal to the reference value, and

(iv) determines whether or not to employ one of the first parallax and the second parallax as the parallax between the matching target block and the block in the second image corresponding to the matching target block, based on magnitudes of the first parallax and the second parallax when both of the first evaluated value and the second evaluated value are greater than or equal to the reference value.

3. The arithmetic processing device according to claim 2 , wherein the processor, in a case in which both of the first evaluated value and the second evaluated value are greater than or equal to the reference value:

(i) employs one of the first parallax and the second parallax as the parallax between the matching target block and the block in the second image corresponding to the matching target block when the magnitudes of the first parallax and the second parallax are equal, the processor, and

(ii) employs neither of the first parallax and the second parallax as the parallax between the matching target block and the block in the second image corresponding to the matching target block when the magnitudes of the first parallax and the second parallax are different.

4. The arithmetic processing device according to claim 1 , wherein

a first range containing the first blocks of the first predetermined number is different from a second range containing the second blocks of the second predetermined number, the first range being defined by the first predetermined number and the first interval, the second range being defined by the second predetermined number and the second interval.

5. The arithmetic processing device according to claim 1 , wherein

the first predetermined number is equal to the second predetermined number, and

the second interval is greater than the first interval.