Method for generating orthophoto image

1. A method for generating an orthophoto image comprising the steps of:

(A) divisionally-capturing a target region from a platform; and

(B) thereafter generating an orthophoto image of an entire target region by applying an orthogonal projection transformation processing to respective divisionally-captured images by use of altitude information of an elevation model of the target region, wherein

in the step of (A) divisionally-capturing the target region,

a plurality of times of image capturing are repeated along a flight direction of the platform flying above the target region with a plurality of cameras as area sensors mounted in a direction perpendicular to the flight direction of the platform flying above the target region

by (a1) rendering viewing angles of the plurality of cameras as being different from each other as well as rendering focal lengths of the plurality of cameras as being different from each other while maintaining a ground resolution of each of the plurality of cameras at a constant reference that is a predetermined ground resolution, along a direction perpendicular to the flight direction, and

by (a2) maintaining the viewing angles of each of the plurality of cameras at the same in the flight direction as well as maintaining the focal lengths of each of the plurality of cameras at the same, along the flight direction, and wherein

in the step of (B) generating the orthophoto image of the entire target region,

(B1) a plurality of elevation models of strip-shaped regions are generated by stereo-matching a plurality of divisionally-captured images, which have been obtained in the same viewing angle and the same focal length by each said camera of said plurality of cameras, in the flight direction by use of tie points thereon with respect to image capturing overlap regions, image capturing position information of each camera of said plurality of cameras, and image capturing orientation information of each camera of said plurality of cameras, where the tie points have a function of associating relative positions of the divisionally-captured images to be matched in the flight direction,

(B2) thereafter the plurality of elevation models of strip-shaped regions are matched in a direction perpendicular to the flight direction among overlap regions corresponding to image capturing overlap regions in the plurality of times of capturing performed in such a manner that the plurality of elevation models adjacent to each other are adjusted with respect to coordinate values of cells thereon corresponding to each other in the overlap region therebetween, and thereby the elevation model of the target region is generated, whereby the cells belonging to said overlap region are adjusted to each other such that the coordinate values of the adjusted cells satisfy a linear expression for adjustment; and

(B3) whereby accuracy of each of the plurality of elevation models of the strip-shaped regions is enhanced while decrease of accuracy of the elevation model of the target region is suppressed.

2. The method for generating an orthophoto image according to claim 1 , wherein

in the divisionally-capturing, focal lengths are set, while ground resolutions are the same, with respect to a plurality of cameras mounted in an airplane so as to be directed at different viewing angles, and wherein

in the generating an elevation model of target region, a plurality of elevation models of strip-shaped regions are generated from respective divisionally-captured images based on image capturing position information and image capturing orientation information of a camera obtained by GPS and IMU mounted in the airplane as well as on positional information of GCP set in a target region and captured, and thereafter, the plurality of elevation models of strip-shaped regions are matched by adjusting respective coordinate values, and wherein

in the generating an orthophoto image, captured images are orthogonally-transformed based on altitude information of the elevation model of the entire target region as well as on the image capturing position information, the image capturing orientation information, and the positional information of the GCP, and thereafter, the captured images are connected by a mosaic processing.

3. The method for generating an orthophoto image according to claim 1 , wherein

in the image capturing, color image capturing is performed and an orthophoto image, which is kept in balance with respect to ground resolution and color depth, is obtained.

4. The method for generating an orthophoto image according to claim 1 , wherein

the divisional capturing includes image capturing in a direction at a nadir of a platform, and wherein

with reference to an elevation model of strip-shaped region based on a captured image in a nadir direction, elevation models of strip-shaped regions based on captured images in other directions are aligned with respect to positions, and thereby elevation models of strip-shaped regions are matched.

5. The method for generating an orthophoto image according to claim 1 , wherein

a mesh is set on overlap regions corresponding to the image capturing overlap region, and a square sum of a difference of coordinate values between the elevation models of strip-shaped regions is minimized in each unit of a plurality of cells in the mesh, and thereby matching is performed.