Panoramic image tone consistency correction method and system

Abstract

The invention proposes a panoramic image tone consistency correction method and system. The method includes the following steps: extracting a content-consistent area from overlapped areas of all adjacent image pairs, and extracting a color corresponding relation between images according to a cumulative probability histogram of the content-consistent area at an equal probability interval; defining a tone transformation model with relatively high flexibility, and optimizing global tone consistency, image contrast and gradient details while designing a global energy function; and through solving model parameters with strict global optimality, applying respective tone correction models to all images to perform tone consistency correction, and outputting a processing result. The panoramic image tone consistency correction method and system utilize the tone corresponding relation of overlapped areas between adjacent images, and can effectively eliminate relatively large tone difference existing among images; the problems of tone overflow, dynamic range narrowing and gradient loss relative to an original image can be suppressed while tone consistency is corrected, a correction model of global joint optimization does not need to define a specific reference image, and the problem of an accumulative error can be effectively reduced.

Description

technical field [0001] The invention belongs to the technical field of image splicing, and in particular relates to a method and system for correcting tone consistency between panoramic and multi-view images. Background technique [0002] Image stitching is a technology that uses multiple images with certain overlapping parts to synthesize high-resolution wide-angle panoramic images. It is currently widely used in panoramic imaging, image rendering, virtual roaming and other fields. In the image stitching process, tone consistency correction is the first step in the tone processing stage (tone correction, stitching line selection, and boundary fusion). The processing effect and efficiency of the two steps. At present, the algorithms for dealing with the difference in hue in image stitching are mainly several classic methods proposed more than ten years ago. On the one hand, the ability to correct the hue is limited, and on the other hand, the corrected image is prone to oth...

AI Technical Summary

Problems solved by technology

[0006] In view of the fact that different images of the same scene may have inconsistent hues due to differences in imaging angles and shoo

Images