A method and device for splicing shelf scene images

Abstract

The invention discloses a shelf scene image splicing method and device. The shelf scene image splicing method comprises the steps of obtaining a plurality of to-be-spliced shelf scene images, whereinall the to-be-spliced goods shelf scene images are sequentially arranged to form an image queue, and an overlapping area exists between every two adjacent to-be-spliced goods shelf scene images; and performing image splicing on the to-be-spliced shelf scene images one by one along the direction from two sides to the middle of the image queue. By implementing the embodiment of the invention, more image contents can be reserved in the image splicing process.

Description

technical field [0001] The invention relates to the technical field of image stitching, in particular to a method and device for stitching images of shelf scenes. Background technique [0002] In the field of fast-moving consumer goods, in order to accurately position and measure the excellence of the market, major fast-moving consumer goods companies need to statistically analyze a large amount of terminal display data. The key prerequisite for effective statistics and analysis of these terminal display data is to first obtain shelf images of offline supermarkets or stores. At present, the collection of shelf images is mainly taken by salesmen running stores. In a real shooting scene, due to limited space or too long shelves, salespersons often need to record complete shelf information by taking multiple images. Therefore, multiple shelf images are spliced ​​effectively and quickly to present a complete shelf scene. However, the existing image stitching technology always...

AI Technical Summary

Problems solved by technology

However, the existing image stitching technology always uses the same direction as the stitching reference direction to stitch all the images. This stitching scheme can easily lead to the loss of part of the image content when stitching multiple images (more than 4 pairs).

For example: Suppose there are several images to be spliced, and the leftmost image is used as the reference image, and the splicing is carried out from left to right with the left as the reference direction, then a cumulative error will be generated, resulting in serious deformation of the rightmost image , resulting in loss of image content after splicing the image on the right

If there is a lot of image content loss in the image stitching in the long shelf scene, it will directly lead to serious errors in the terminal display data of the user's final statistics.

Images