Panoramic image stitching method and system

Abstract

The embodiment of the invention provides a panoramic image splicing method and system, and the method comprises the steps: carrying out the registration of target scene point cloud data obtained through the scanning of a target scene through a multi-line laser radar corresponding to a panoramic camera, obtaining a second type of single images, and determining the overlapping region of the second type of single images corresponding to every two adjacent lenses; performing dynamic partitioning on the overlapped area through the first type of distance information carried in the second type of single image corresponding to every two adjacent lenses, so that each partition obtained through dynamic partitioning corresponds to one second type of distance information; and finally, performing panoramic image splicing on all the first type of single images according to the second type of distance information corresponding to each partition and the distance splicing template corresponding to thesecond type of distance information. Partitions with different distance information in an overlapped area are spliced by adopting different distance splicing templates, so that the problem of ghostingcaused by panoramic image splicing when a plurality of different distances exist in the overlapped area and a target is put in is solved, and the splicing accuracy is ensured.

Description

technical field [0001] The present invention relates to the field of image processing technology and the field of multi-line lidar ranging technology, and more specifically, to a panoramic image stitching method and system. Background technique [0002] At present, to achieve panoramic image stitching, feature-based matching methods such as Scale-invariant featuretransform (SIFT) and Speed-Up Robust Features (SURF) are mostly used, and the overlapping parts of two images To perform feature matching, these feature-based matching methods require more feature information in the two images. But when there is little or no feature information in the two images, a matching error will occur. [0003] In the prior art, when stitching panoramic images, lens calibration can be performed according to the geometric relationship between the lenses, a transformation matrix can be generated, and stitching can be performed after conversion. However, this method does not consider multiple o...

AI Technical Summary

Problems solved by technology

However, this method does not consider multiple objects with different distances in the image. If multiple objects with different distances appear in the overlapping part of the image at the same time, and the panoramic camera has a large lens distortion, stitching errors will also occur, and Objects at different distances will produce different degrees of ghosting and other problems, resulting in inaccurate stitching, unreal panoramic images, and affecting the visual effects and research applications of panoramic images.

Images