An automatic generation method of mosaic puzzle based on visual features

Abstract

The invention belongs to the technical field of image processing, and discloses an automatic generation method of mosaic puzzles based on visual features, which splices a large number of small images into a pair of large images. Compared with the current semi-automatic production method, the automatic generation method is based on the underlying characteristics of the image, and finds an optimal image stitching sequence from a limited set of small image materials, thereby reducing the abrupt noise blocks in the mosaic; The image overlay method avoids the loss of local details of the large image at the macro level. The method specifically includes: rectangular segmentation of large images; image bottom-level feature extraction; small image screening based on image bottom-level features; adaptive image superposition based on signal-to-noise ratio. From the effect of the jigsaw puzzle, the method of the present invention has a better visual effect compared with the current jigsaw puzzle method, and the error with the original picture is smaller; the visual effect of manual adjustment is achieved, and a fully automatic mosaic jigsaw puzzle is realized.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to an automatic generation method of a mosaic puzzle based on visual features. Background technique [0002] The mosaic puzzle technology is to automatically assemble a large number of images into a mosaic image with a montage effect. The mosaic puzzle with montage effect refers to the splicing and combination of many small images to generate another large image. From a macro perspective, it is a large image; and in detail, each image block is also a separate image. . A single image has limited ability to express, and lacks the ability to describe time and context, while the mosaic puzzle with montage effect can express richer connotations, and has a wide range of applications in film art, visual art, and interior design. [0003] At present, the automatic generation method of mosaic image with montage effect is mainly generated through operations such as imag...

AI Technical Summary

Problems solved by technology

[0008] There are many abrupt noise blocks in the existing automatically generated large mosaic images; since the position where the small image is superimposed on the large image is randomly selected, when combined to the large image, if the background difference is large, the superposition of the small image Instead of producing quantitative visual effects, it becomes the noise of the large image. The reason for these small noise images is the selection of small image material sets on the one hand; An optimal combination method;

[0009] The local information of the large image is lost; when the mosaic image is generated, the large image and the small image need to be overlaid; in the existing methods, the same overlay coefficient is used for operation; although this method can be manually adjusted to a relatively good effect, but In some details, the details of the large picture will inevitably be lost

Images