Image data set construction method for morphological prior constrained oracle bone text detection

An image data set and construction method technology, which is applied in the field of image data set construction for oracle bone inscription detection with morphological prior constraints, can solve the problems of limited and difficult oracle bone inscription rubbing image data sets, lack of annotations in oracle bone inscription detection data sets, etc. , to achieve the effect of improving labeling efficiency and eliminating the impact of labeling accuracy

Active Publication Date: 2021-09-07
LIAONING NORMAL UNIVERSITY
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to solve the above-mentioned technical problems existing in the prior art, and provides a method that can effectively solve the problems of limited oracle bone inscription rubbing image data sets, lack of oracle bone character detection data sets and difficulty in labeling, and can reflect real oracle bone rubbing images. The characteristics of noise and inherent texture, high construction efficiency, strong scalability, automatic labeling ability, high labeling accuracy, and morphological prior constraints. An image dataset construction method for oracle bone inscription detection

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image data set construction method for morphological prior constrained oracle bone text detection
  • Image data set construction method for morphological prior constrained oracle bone text detection
  • Image data set construction method for morphological prior constrained oracle bone text detection

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] The image data set construction method of the oracle bone inscription detection of the morphological prior constraints of the present invention is carried out according to the following steps;

[0041] Step 1. input the rubbing images of N oracle bone inscriptions characters to form a real rubbing image data set RI, in the present embodiment, make N=39476;

[0042] Step 2. Segment and extract clear M from the real rubbing image dataset RI by manual interaction. 1 A piece of oracle bone inscription single-character image forms a font image data set ST. In this embodiment, let M 1 =22524;

[0043] Step 3. Segment and extract clear M from the real rubbing image dataset RI by manual interaction. 2 Speckle and noise images form the interference element data set SN, in this embodiment, let M 2 =5000;

[0044] Step 3.1 Segment and extract M from RI 21 The image of point-like noise, this type of noise is often caused by reasons such as rough surface texture of oracle bone ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses an image data set construction method for morphological prior constrained oracle bone text detection. The method comprises the following steps of: extracting noise-free single character images and interference element images containing spot noise, platy stripes and inherent textures from a real oracle bone rubbing image, and respectively forming a single character font image data set and an interference element data set; extracting pixels of an effective character font area by using a binary mask, and generating a noise-free oracle bone rubbing simulation image through random rotation, scaling and bit operation based on the binary mask by taking oracle bone character morphological characteristics as constraints; positioning a bounding box of a character area through histogram scanning, and generating an XML annotation file conforming to the VOC2007 standard; and through random rotation, scaling and binary mask modes, superposing interference elements and single character images, thereby constructing an image data set for oracle bone divination character detection.

Description

technical field [0001] The present invention relates to the intersection field of digital image processing and ancient character information processing, especially a kind of can effectively solve the problems of limited oracle bone inscription rubbing image data set, lack of oracle bone character detection data set and difficulty in labeling, and can reflect the real oracle bone rubbing image of text, Based on the characteristics of noise and inherent texture, it has high construction efficiency, strong scalability, automatic labeling ability, high labeling accuracy, and an image dataset construction method for oracle bone character detection oriented to morphological prior constraints. Background technique [0002] Oracle bone inscription detection is one of the important basic research fields of computational oracle bone science. Its fundamental purpose is to automatically locate the area position of oracle bone inscriptions on oracle bone rubbing images with the help of co...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/155G06T7/62G06T7/70G06T5/40
CPCG06T7/155G06T7/62G06T7/70G06T5/40
Inventor 宋传鸣何熠辉王一琦洪飏王相海
Owner LIAONING NORMAL UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap