Camera parameter set calculation method, recording medium, and camera parameter set calculation apparatus

1. A camera parameter set calculation method comprising:

(a1) acquiring a first image captured by a first camera and a second image captured by a second camera;

(a2) acquiring a first camera parameter set including one or more camera parameters of the first camera and a second camera parameter set including one or more camera parameters of the second camera;

(a3) calculating three-dimensional coordinate sets on an object on a basis of the first image, the second image, the first camera parameter set, and the second camera parameter set, the first image including an image of the object, the second image including an image of the object;

(a4) determining first pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the first image on a basis of the first camera parameter set and determining second pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the second image on a basis of the second camera parameter set;

(a5) calculating an evaluation value on a basis of pixel values at the first pixel coordinate pairs in the first image and pixel values at the second pixel coordinate pairs in the second image;

(a6) updating the first camera parameter set and the second camera parameter set on a basis of the evaluation value; and

(a7) outputting the updated first camera parameter set and second camera parameter set,

wherein at least one of processes (a1) to (a7) is performed by a processor,

the three-dimensional coordinate sets are represented by (x1, y1, z1) . . . (xi, yi, zi) . . . (xn, yn, zn), the first pixel coordinate pairs are represented by (u11, v11) . . . (u1i, v1i) . . . (u1n, v1n), and the second pixel coordinate pairs are represented by (u21, v21) . . . (u2i, v2i) . . . (u2n, v2n), where 1<i≤n, and i and n are natural numbers,

the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u1i, v1i), the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u2i, v2i), the pixel value at the pixel coordinate pair (u1i, v1i) is represented by I1(u1i, v1i), and the pixel value at the pixel coordinate pair (u2i, v2i) is represented by I2(u2i, v2i), and

in process (a5), the evaluation value is calculated on a basis of a difference Di between the pixel values I1(u1i, v1i) and I2(u2i, v2i).

2. The camera parameter set calculation method according to claim 1 , wherein in process (a4), the first pixel coordinate pairs are determined by coordinate-transforming each of the three-dimensional coordinate sets using one or more camera parameters included in the first camera parameter set, and the second pixel coordinate pairs are determined by coordinate-transforming each of the three-dimensional coordinate sets using one or more camera parameters included in the second camera parameter set.

3. The camera parameter set calculation method according to claim 1 , wherein in process (a5), the evaluation value is calculated on a basis of Σ i=1 n Di.

4. The camera parameter set calculation method according to claim 1 , wherein in process (a5), the evaluation value is calculated on a basis of Σ i=1 n |Di|.

5. The camera parameter set calculation method according to claim 1 , wherein in process (a5), the evaluation value is calculated on a basis of Σ i=1 n (Di×Di).

6. The camera parameter set calculation method according to claim 1 , wherein (b1) information about detected temperature is acquired from a temperature sensor before processes (a1) to (a7) are performed, and

wherein (b2) processes (a1) to (a7) are started on a basis of the acquired information about detected temperature.

7. The camera parameter set calculation method according to claim 1 , wherein (c1) information about detected impact is acquired from an impact sensor before processes (a1) to (a7) are performed, and

wherein (c2) processes (a1) to (a7) are started on a basis of the acquired information about detected impact.

8. The camera parameter set calculation method according to claim 1 , wherein (d1) information about time is acquired from a timer before processes (a1) to (a7) are performed, and

wherein (d2) processes (a1) to (a7) are started on a basis of the acquired information about time.

9. A camera parameter set calculation method comprising:

(a1) acquiring a first image captured by a first camera and a second image captured by a second camera;

(a2) acquiring a first camera parameter set including one or more camera parameters of the first camera and a second camera parameter set including one or more camera parameters of the second camera;

(a3) repeating estimation of a pair of corresponding points that are a first corresponding point in the first image and a second corresponding point in the second image, to determine which points in the first image and in the second image indicates images of an identical point on the object, based on a pixel value at the first corresponding point and a pixel value at the second corresponding point, thereby pairs of corresponding points being provided;

(a4) calculating three-dimensional coordinate sets each corresponding to the pairs of corresponding points on a basis of the first camera parameter set and the second camera parameter set;

(a5) determining first pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the first image on a basis of the first camera parameter set and determining second pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the second image on a basis of the second camera parameter set;

(a6) calculating an evaluation value on a basis of pixel values at the first pixel coordinate pairs in the first image and pixel values at the second pixel coordinate pairs in the second image;

(a7) updating the first camera parameter set and the second camera parameter set on a basis of the evaluation value; and

(a8) outputting the updated first camera parameter set and second camera parameter set,

wherein at least one of processes (a1) to (a8) is performed by a processor,

the three-dimensional coordinate sets are represented by (x1, y1, z1) . . . (xi, yi, zi) . . . (xn, yn, zn), the first pixel coordinate pairs are represented by (u11, v11) . . . (u1i, v1i) . . . (u1n, vin), and the second pixel coordinate pairs are represented by (u21, v21) . . . (u2i, v2i) . . . (u2n, v2n), where 1<i≤n, and i and n are natural numbers,

the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u1i, v1i), the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u2i, v2i), the pixel value at the pixel coordinate pair (u1i, v1i) is represented by I1(u1i, v1i), and the pixel value at the pixel coordinate pair (u2i, v2i) is represented by I2(u2i, v2i), and

in process (a6), the evaluation value is calculated on a basis of a difference Di between the pixel values I1(u1i, v1i) and I2(u2i, v2i).

10. A non-transitory computer-readable recording medium storing a control program executable by an apparatus including a processor, the control program comprising code to perform:

(a1) acquiring a first image captured by a first camera and a second image captured by a second camera;

(a2) acquiring a first camera parameter set including one or more camera parameters of the first camera and a second camera parameter set including one or more camera parameters of the second camera;

(a3) calculating three-dimensional coordinate sets on an object on a basis of the first image, the second image, the first camera parameter set, and the second camera parameter set, the first image including an image of the object, the second image including an image of the object;

(a4) determining first pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the first image on a basis of the first camera parameter set and determining second pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the second image on a basis of the second camera parameter set;

(a5) calculating an evaluation value on a basis of pixel values at the first pixel coordinate pairs in the first image and pixel values at the second pixel coordinate pairs in the second image;

(a6) updating the first camera parameter set and the second camera parameter set on a basis of the evaluation value; and

(a7) outputting the updated first camera parameter set and second camera parameter set,

wherein the three-dimensional coordinate sets are represented by (x1, y1, z1) . . . (xi, yi, zi) . . . (xn, yn, zn), the first pixel coordinate pairs are represented by (u11, v11) . . . (u1i, v1i) . . . (u1n, vin), and the second pixel coordinate pairs are represented by (u21, v21) . . . (u2i, v2i) . . . (u2n, v2n), where 1<i≤n, and i and n are natural numbers,

the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u1i, v1i), the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u2i, v2i), the pixel value at the pixel coordinate pair (u1i, v1i) is represented by I1(u1i, v1i), and the pixel value at the pixel coordinate pair (u2i, v2i) is represented by I2(u2i, v2i), and

in process (a5), the evaluation value is calculated on a basis of a difference Di between the pixel values I1(u1i, v1i) and I2(u2i, v2i).

11. A camera parameter set calculation apparatus comprising a processing circuit that performs the following operations:

(a1) acquiring a first image captured by a first camera and a second image captured by a second camera;

(a2) acquiring a first camera parameter set including one or more camera parameters of the first camera and a second camera parameter set including one or more camera parameters of the second camera;

(a3) calculating three-dimensional coordinate sets on an object on a basis of the first image; the second image; the first camera parameter set; and the second camera parameter set, the first image including an image of the object, the second image including an image of the object;

(a4) determining first pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the first image on a basis of the first camera parameter set and determining second pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the second image on a basis of the second camera parameter set;

(a5) calculating an evaluation value on a basis of pixel values at the first pixel coordinate pairs in the first image and pixel values at the second pixel coordinate pairs in the second image;

(a6) updating the first camera parameter set and the second camera parameter set on a basis of the evaluation value; and

(a7) outputting the updated first camera parameter set and second camera parameter set,

wherein the three-dimensional coordinate sets are represented by (x1, y1, z1) . . . (xi, yi, zi) . . . (xn, yn, zn), the first pixel coordinate pairs are represented by (u11, v11) . . . (u1i, v1i) . . . (u1n, vin), and the second pixel coordinate pairs are represented by (u21, v21) . . . (u2i, v2i) . . . (u2n, v2n), where 1<i≤n, and i and n are natural numbers,

the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u1i, v1i), the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u2i, v2i), the pixel value at the pixel coordinate pair (u1i, v1i) is represented by I1(u1i, v1i), and the pixel value at the pixel coordinate pair (u2i, v2i) is represented by I2(u2i, v2i), and

in process (a5), the evaluation value is calculated on a basis of a difference Di between the pixel values I1(u1i, v1i) and I2(u2i, v2i).

12. A camera parameter set calculation apparatus comprising a processing circuit that performs the following operations:

(a1) acquiring a first image captured by a first camera and a second image captured by a second camera;

(a2) acquiring a first camera parameter set including one or more camera parameters of the first camera and a second camera parameter set including one or more camera parameters of the second camera;

(a3) repeating estimation of a pair of corresponding points that are a first corresponding point in the first image and a second corresponding point in the second image, to determine which points in the first image and in the second image indicates images of an identical point on the object, based on a pixel value at the first corresponding point and a pixel value at the second corresponding point, thereby pairs of corresponding points being provided;

(a4) calculating three-dimensional coordinate sets each corresponding to the pairs of corresponding points on a basis of the first camera parameter set and the second camera parameter set;

(a5) determining first pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the first image on a basis of the first camera parameter set and determining second pixel coordinate pairs obtained by projecting the three-dimensional coordinate sets onto the second image on a basis of the second camera parameter set;

(a6) calculating an evaluation value on a basis of pixel values at the first pixel coordinate pairs in the first image and pixel values at the second pixel coordinate pairs in the second image;

(a7) updating the first camera parameter set and the second camera parameter set on a basis of the evaluation value; and

(a8) outputting the updated first camera parameter set and second camera parameter set,

wherein the three-dimensional coordinate sets are represented by (x1, y1, z1) . . . (xi, yi, zi) . . . (xn, yn, zn), the first pixel coordinate pairs are represented by (u11, v11) . . . (u1i, v1i) . . . (u1n, vin), and the second pixel coordinate pairs are represented by (u21, v21) . . . (u2i, v2i) . . . (u2n, v2n), where 1<i≤n, and i and n are natural numbers,

the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u1i, v1i), the three-dimensional coordinate set (xi, yi, zi) corresponds to the pixel coordinate pair (u2i, v2i), the pixel value at the pixel coordinate pair (u1i, v1i) is represented by I1(u1i, v1i), and the pixel value at the pixel coordinate pair (u2i, v2i) is represented by I2(u2i, v2i), and

in process (a6), the evaluation value is calculated on a basis of a difference Di between the pixel values I1(u1i, v1i) and I2(u2i, v2i).