Calligraphy character smallest bounding box extraction method based on wavelet smoothing

Abstract

The invention discloses a calligraphy character smallest bounding box extraction method based on wavelet smoothing. The method concretely comprises the following steps: first of all, inputting inscription images, and obtaining binary calligraphy images after preprocessing; performing column integration on the binary calligraphy images, and segmenting an initial position and an ending position of each column; performing row integration on each column, introducing the wavelet smoothing for performing the wavelet smoothing on row integration signals to obtain smooth waveforms, and segmenting rows on the smooth waveforms; merging fractured characters; performing column integration again on each character firstly, and cutting accurate left edges and right edges; afterwards, recording upper, lower, left and right edges of each character; and finally, reserving cut single-character images. The calligraphy character smallest bounding box extraction method based on the wavelet smoothing can realize cutting of the inscription images in horizontal rows and vertical columns and can effectively process noise.

Description

technical field [0001] The invention belongs to the technical field of image processing methods, and in particular relates to a method for extracting minimum bounding boxes of calligraphy characters based on wavelet smoothing. Background technique [0002] In recent years, with the rapid development and popularization of computer technology, more computer technology has begun to be used in the traditional calligraphy field. In the process of processing the scanned rubbing image, it is necessary to extract a single Chinese character from the rubbing image, and find the minimum bounding box of the character. [0003] The so-called minimum bounding box refers to the smallest rectangular frame surrounding a character, which is realized by the method of horizontal and vertical integration. The upper and lower edges of each character, and finally do a column integration to divide the left and right edges of each word. How to accurately and simply realize the row and column cuttin...

AI Technical Summary

Problems solved by technology

[0003] The so-called minimum bounding box refers to the smallest rectangular frame surrounding a character, which is realized by the method of horizontal and vertical integration. The upper and lower edges of each character, and finally do a column integration to divide the left and right edges of each word. How to accurately and simply realize the row and column cutting in this method is difficult. The traditional row and column integration method only performs column integration. In fact, there are two row integration methods. This step has the following disadvantages in actual operation: (1) Due to the presence of noise such as stele post image defacement, there are glitches in the line integral waveform, which makes it difficult to judge the beginning and end of the character, resulting in the cutting of the upper and lower edges of the above method is not accurate enough. The smallest bounding box with precise characters; (2) Some Chinese characters may be cut in half, etc.

Images