Rapid panoramic stitching method and device suitable for space complex environment

Abstract

The invention provides a rapid panoramic stitching technology suitable for a space complex environment, and the method at least comprises the following steps: S101, carrying out the calibration test of a camera, and extracting an angular point on a calibration image photographed by an effective view field; S102, obtaining an internal parameter of each camera in a single-target calibration manner;S103, performing double-target calibration to obtain external parameters between the adjacent cameras; S104, performing spherical back projection to form a 360-degree*180-degree panoramic plane image,selecting one path of camera as a basic reference coordinate system, completing image coordinate conversion of the other three paths, and realizing indexing of image pixel points under the same coordinate system; s105, adopting a gradual-in gradual-out algorithm to eliminate splicing seams, and carrying out smoothing processing on a large parallax pixel area; and S106, making the pixel point indexes of the four images corresponding to the panoramic image and the fusion coefficient of the stitching area into a LUT table used by the FPGA.

Description

technical field [0001] The invention relates to the technical field of video image synthesis, in particular to a fast panorama stitching method and device suitable for complex space environments. Background technique [0002] Panoramic vision refers to the acquisition of all visual information in a three-dimensional space larger than the hemispherical field of view (360° x 180°) at one time. Achieving panoramic vision requires a special vision sensor system. Panoramic vision, as the most direct way to obtain three-dimensional spatial information, has been widely used in many fields, especially for fields and industries that make decisions through visual information, such as civil, military, and aerospace. [0003] At present, the methods of obtaining panoramic visual images are: (1) the method of ordinary visual sensor + rotating pan-tilt, the field of view of ordinary visual sensor is limited, and the field of view is enlarged by the rotation of the pan-tilt; (2) the metho...

AI Technical Summary

Problems solved by technology

[0003] At present, the methods of obtaining panoramic visual images are: (1) the method of ordinary visual sensor + rotating pan-tilt, the field of view of ordinary visual sensor is limited, and the field of view is enlarged by the rotation of the pan-tilt; (2) the method of compound eye technology + image stitching, using Multiple visual sensors acquire visual images at different angles of the field of view at the same time, and then realize the seamless splicing of the images; (3) Using fisheye imaging technology, the fish's eyes can observe the field of view close to the hemisphere at one time. According to the fisheye imaging Principle The specially manufactured fisheye lens is composed of multiple groups of lenses, the imaging principle is complex, and the price is relatively expensive; (4) The method of using a convex mirror + ordinary visual sensor

[0004] The cost of the existing technology is too high, the imaging in the complex space environment is not stable enough, and the stitching effect is not ideal

Images