Image processing apparatus and method that synthesizes an all-round image of a vehicle's surroundings

• US 9,633,266 B2
• Filed: 09/22/2016
• Issued: 04/25/2017
• Est. Priority Date: 10/25/2012
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Shimizu, Seiya
• Primary Examiner(s): GEROLEO, FRANCIS
• Application Number:
  US15/273,075
• Publication Number:
  US 20170011267A1
• Time in Patent Office: 215 Days
• Field of Search: 348148
• CPC Class Codes: G06T 15/00 3D [Three Dimensional] image rendering
• CPC Class Codes: G06T 15/205 Image-based rendering
• CPC Class Codes: G06T 3/4038 Image mosaicing, e.g. composing plane images from plane sub-images
• CPC Class Codes: G06T 7/00 Image analysis
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle

Abstract

An image processing apparatus synthesize a synthesized image by taking pixel values of pixels in a synthesized image corresponding to each point on the three dimensional projection plane as viewed from a specific viewpoint position, as pixel values of corresponding pixels of the first image based on the first correspondence relationship, and taking pixel values of each pixel in the synthesized image corresponding to pixels identified in the first image as being pixels representing a solid object, as pixel values of corresponding pixels of the second image based on the second correspondence relationship.

Claims

1. An image processing apparatus, comprising:

processor; and

a memory storing instructions, which, when executed by the processor, perform a procedure, the procedure including;

(a) acquiring image data representing successively captured images of a moving body'"'"'s surroundings, and acquiring three-dimensional (3D) position data representing 3D positions of a plurality of points on a solid object present in the moving body'"'"'s surroundings as expressed in a moving body coordinate system set for the moving body;

(b) calculating a movement amount that the moving body has moved during an interval from a first time to a second time;

(c) identifying pixels representing a solid object in a first image captured at the first time based on the 3D position data acquired by (a);

(f) expressing a position of the moving body at the second time based on the movement amount calculated by (b) in the moving body coordinate system at the first time, and infilling pixel values of pixels in the first image identified by (c) as being pixels representing a solid object with corresponding pixel values of pixels of a second image captured at the second time, employing a virtual viewpoint projection that projects a captured image onto a virtual viewpoint projection plane that has a shape configured by a combination of a hemisphere of infinite radius with the position of the moving body as its origin, and a horizontal plane in which the moving body is present, and the virtual viewpoint projection re-projects points projected onto the virtual viewpoint projection plane onto a three dimensional projection plane with reference to a virtual viewpoint position;

(g) creating, by the virtual viewpoint projection, a first correspondence relationship indicating correspondence between each point on the three dimensional projection plane determined so as to contain the moving body therein at the first time, and positions of each pixel of the first image; and

(h) synthesizing a synthesized image by taking pixel values of each pixel on a synthesized image corresponding to each point on the three dimensional projection plane as viewed from a specific viewpoint position, as pixel values of corresponding pixels of the first image after infilling in (f) based on the first correspondence relationship.

2. An image processing method to be executed by a processor, comprising:

(a) acquiring image data representing successively captured images of a moving body'"'"'s surroundings, and acquiring three-dimensional (3D) position data representing 3D positions of a plurality of points on a solid object present in the moving body'"'"'s surroundings as expressed in a moving body coordinate system set for the moving body;

(b) calculating a movement amount that the moving body has moved during an interval from a first time to a second time;

(c) identifying pixels representing a solid object in a first image captured at the first time based on the 3D position data acquired by (a);

(f) expressing a position of the moving body at the second time based on the movement amount calculated by (b) in the moving body coordinate system at the first time, and infilling pixel values of pixels in the first image identified by (c) as being pixels representing a solid object with corresponding pixel values of pixels of a second image captured at the second time;

employing a virtual viewpoint projection that projects a captured image onto a virtual viewpoint projection plane that has a shape configured by a combination of a hemisphere of infinite radius with the position of the moving body as its origin, and a horizontal plane in which the moving body is present, and the virtual viewpoint projection re-projects points projected onto the virtual viewpoint projection plane onto a three dimensional projection plane with reference to a virtual viewpoint position;

(g) creating, by the virtual viewpoint projection, a first correspondence relationship indicating correspondence between each point on the three dimensional projection plane determined so as to contain the moving body therein at the first time and positions of each pixel of the first image; and

(h) synthesizing a synthesized image by taking pixel values of each pixel on a synthesized image corresponding to each point on the three dimensional projection plane as viewed from a specific viewpoint position, as pixel values of corresponding pixels of the first image after in-filling in (f) based on the first correspondence relationship.

3. A non-transitory computer-readable recording medium having stored therein a program for causing a computer to execute an image processing process comprising:

(a) acquiring image data representing successively captured images of a moving body'"'"'s surroundings, and acquiring three-dimensional (3D) position data representing 3D positions of a plurality of points on a solid object present in the moving body'"'"'s surroundings as expressed in a moving body coordinate system set for the moving body;

(b) calculating a movement amount that the moving body has moved during an interval from a first time to a second time;

(c) identifying pixels representing a solid object in a first image captured at the first time based on the 3D position data acquired by (a);

(f) expressing a position of the moving body at the second time based on the movement amount calculated by (b) in the moving body coordinate system at the first time, and infilling pixel values of pixels in the first image identified by (c) as being pixels representing a solid object with corresponding pixel values of pixels of a second image captured at the second time;

employing a virtual viewpoint projection that projects a captured image onto a virtual viewpoint projection plane that has a shape configured by a combination of a hemisphere of infinite radius with the position of the moving body as its origin, and a horizontal plane in which the moving body is present, and the virtual viewpoint projection re-projects points projected onto the virtual viewpoint projection plane onto a three dimensional projection plane with reference to a virtual viewpoint position;

(g) creating, by the virtual viewpoint projection, a first correspondence relationship indicating correspondence between each point on the three dimensional projection plane determined so as to contain the moving body therein at the first time and positions of each pixel of the first image; and

(h) synthesizing a synthesized image by taking pixel values of each pixel on a synthesized image corresponding to each point on the three dimensional projection plane as viewed from a specific viewpoint position, as pixel values of corresponding pixels of the first image after infilling in (f) based on the first correspondence relationship.