Form content extracting method based on registration of local feature points of image

Abstract

The present invention proposes a form content extracting method based on registration of local feature points of an image. The method comprises the following steps of: extracting SIFT feature points of an input image; solving a set of matched feature point pairs of each local image block of the input image and a corresponding reference image; calculating transformation matrices of each local image block of the reference image according to the matched feature point pairs; obtaining each corresponding local image block of the input image according to the transformation matrices; establishing an image model by identifying the local image block as a node and a geometric relationship among the local image blocks as a side; matching each local image block of the input image and each local image block of the reference image by using image matching calculation to obtain a set of matched image blocks; obtaining a global transformation matrix according to the set of matched feature point pairs in the set of matched image blocks; transforming an image by using the global transformation matrix; subtracting the transformed input image and the reference image to obtain a difference of a corresponding pixel; and and binarizing the subtracted input images to obtain a form content.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a form content extraction method based on image local feature point registration. Background technique [0002] In recent years, with the rapid development of e-commerce, the number of express parcels has increased dramatically. The form in the express package contains a lot of information such as the sender of the package, the recipient, the address of the recipient, and the mailing item. This information is very important for sorting mail, information extraction, and security inspection. To identify the information on the package form, it is first necessary to extract the content entered by the user, and then perform further work such as identification. [0003] Due to differences in lighting, image acquisition equipment, and package size, the processing of the package form image is more complicated in terms of resolution, brightness, and tilt angle. Therefore, there is an urg...

AI Technical Summary

Problems solved by technology

[0003] Due to differences in lighting, image acquisition equipment, and package size, the processing of package form images is more complicated in terms of resolution, brightness, and tilt angles.

Images