Panoramic image fusion method, system and image processing equipment

Abstract

The invention is suitable for the technical field of image processing, providing a panoramic image fusion method. The panoramic picture fusion method comprises the following steps: calculating the projection transformation parameters of two images A and B which have certain coincident viewing field to a panoramic viewing field; projecting and transforming the images into a unified panoramic viewing field; calculating the public viewing field overlapping region of the two images in the panoramic viewing field; calculating the dissimilar viewing field boundary of the public viewing field overlapping region; calculating distance from the pixel points of the public viewing field overlapping region to the dissimilar viewing field boundary; according to the distance of the dissimilar viewing field boundary, calculating fusion coefficients; and according to a fusion coefficient table, fusing the images of the public viewing field area. The invention solves the problem that the existing panoramic image fusion method can not better handle the problem of seamless fusion when the public viewing field overlapping region is in any shape, and also solves the problem that the existing fusion algorithms are not unified because of different shapes of the public viewing field overlapping region.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a panoramic image fusion method, system and image / video processing equipment. Background technique [0002] At present, the generation of panoramic images generally adopts two types of methods: the first type is to use wide-angle lenses or panoramic reflection optical lenses to generate panoramic images; Large-field panoramic images of image content from various perspectives. [0003] In the second type of method, due to the differences in the field of view, direction, brightness, color, etc. between the images of multiple viewing angles acquired by the camera, it generally takes two steps to stitch these images into a seamless image. panoramic image. The first step is to projectively transform each image to transform them into a unified field of view; the second step is to perform color and brightness consistency on the transformed images in their common f...

AI Technical Summary

Problems solved by technology

Although the latter method can overcome the shortcomings of the previous method, it can only calculate the case where the overlapping area is a quadrilateral, which limits its scope of use.

[0010] When the existing fade-in and fade-out fusion technology generates a panoramic image, there is a problem that the overlapping area of ​​the common field of view cannot be effectively fused when the overlapping area is of any geometric shape, and the fusion algorithm caused by the different shapes of the overlapping area of ​​the common field of view inconsistency problem

Images