Method and apparatus for generating learning data for neural network

A method for generating learning data for a neural network may comprise: a step of generating a license plate image by combining a background image, a frame image, and a text image; a step of generating a conversion image by performing at least one conversion among a geometric conversion and filter...

Full description

Saved in:
Bibliographic Details
Main Authors NOH GYEONG GI, KIM DONG GEON
Format Patent
LanguageEnglish
Korean
Published 07.02.2023
Subjects
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:A method for generating learning data for a neural network may comprise: a step of generating a license plate image by combining a background image, a frame image, and a text image; a step of generating a conversion image by performing at least one conversion among a geometric conversion and filter conversion for the license plate image; a step of setting a text corresponding to the text image as the target data corresponding to the conversion image; and a step of generating learning data comprising the conversion image and target data. Therefore, the present invention is capable of efficiently generating learning data. 뉴럴 네트워크를 위한 학습 데이터 생성 방법은 배경 이미지, 프레임 이미지 및 텍스트 이미지를 결합함으로써 번호판 이미지를 생성하는 단계, 번호판 이미지에 대해 기하 변환 및 필터 변환 중 적어도 하나의 변환을 수행함으로써 변환 이미지를 생성하는 단계, 텍스트 이미지에 대응되는 텍스트를 변환 이미지에 대응되는 목표 데이터로 설정하는 단계 및 변환 이미지 및 목표 데이터를 포함하는 학습 데이터를 생성하는 단계를 포함할 수 있다.
Bibliography:Application Number: KR20210100346