Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Oracle carved text detection method combining local prior features and deep convolution features

A technology of deep convolution and text detection, applied in neural learning methods, computer components, character and pattern recognition, etc., can solve problems such as inability to achieve accurate extraction of oracle bone characters

Active Publication Date: 2020-06-19
LIAONING NORMAL UNIVERSITY
View PDF4 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the above two methods can only locate the location of the rubbings in the oracle bone rubbing image, but cannot achieve accurate extraction of oracle bone characters
[0006] To sum up, there is currently no one that can effectively resist point noise, patchy markings and inherent texture interference on tortoise shells and animal bones. It has high accuracy, good robustness, fast processing speed, and self-adaptive ability. Oracle bone script automatic detection method

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
  • Oracle carved text detection method combining local prior features and deep convolution features
  • Oracle carved text detection method combining local prior features and deep convolution features
  • Oracle carved text detection method combining local prior features and deep convolution features

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0101] The oracle bone inscription character detection method of the present invention in conjunction with local prior feature and deep convolution feature, carries out according to the following steps;

[0102] Step 1. Establish and initialize a deep convolutional neural network N for detection of oracle bone inscriptions obc , the N obc Contains 1 subnetwork N for feature extraction feature , 1 sub-network N for oracle region proposals rpn , 1 sub-network N for dimensionality reduction of oracle bone character area features dimand 1 subnetwork N for region classification cls ;

[0103] Step 1.1 Establish and initialize subnetwork N feature , containing 6 sets of convolutional layers, namely Conv0, Conv1, Conv2, Conv3, Conv4 and Conv5;

[0104] The Conv0 includes 1 layer of convolution operation and 5 preset convolution kernels defined by formula (1), each convolution kernel performs convolution operation with a step size of 1 pixel;

[0105]

[0106] The Conv1 incl...

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 oracle carved text detection method combining local prior features and deep convolution features, and belongs to the crossing field of digital image processing and ancient text information processing. The method comprises the following steps: firstly, inputting a to-be-processed oracle carved image I, converting the color space of the to-be-processed oracle carved imageI from RGB to HSV, binarizing a V component of the to-be-processed oracle carved image I by adopting a maximum between-cluster variance method, and obtaining a rough denoising result of the image I byutilizing a multi-condition connected region filling algorithm; secondly, analyzing deep convolution features of the image I through a feature extraction sub-network containing six groups of convolution layers; secondly, selecting a region of interest by utilizing the region suggestion sub-network containing three groups of convolution layers, and calculating a score of each region of interest; and finally, determining an area of the oracle characters by using the feature dimension reduction sub-network and the area classification sub-network, and removing overlapped candidate boxes by usinga non-maximum suppression algorithm to generate a result image with a box selection result.

Description

technical field [0001] The present invention relates to the intersection field of digital image processing and ancient text information processing, especially a kind of can effectively resist the interference of dot noise, flaky markings and inherent texture on tortoise shell and animal bone, high accuracy, good robustness, processing A fast and self-adaptive oracle bone inscription text detection method that combines local prior features and deep convolution features. Background technique [0002] Oracle bone inscriptions are written in the late Shang Dynasty using tortoise shells and animal bones for recording and divination. As the earliest mature writing system discovered in my country so far, it has extremely important protection value, research value and lofty symbolic significance. Oracle bone inscriptions are not only the earliest traceable source of Chinese language, culture, and history, but also the source of Chinese characters and the root of Chinese excellent t...

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): G06K9/32G06K9/62G06N3/04G06N3/08
CPCG06N3/084G06V20/62G06V30/287G06N3/047G06N3/045G06F18/241G06F18/2415Y02D10/00
Inventor 宋传鸣何熠辉洪飏王相海
Owner LIAONING NORMAL UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products