摘要
针对手动提取图像细节时局部的缺失问题,文章提出一种基于U^(2)Net神经网络算法的图像外轮廓提取技术,并将其运用于河南博物院文创产品的数字化分析设计上。此方法从编码器提取妇好鴞尊图像特征,利用解码器基于这些特征生成分割图,大大节约设计师手动提取文物元素造型的时间效率。
To solve the problem of partial missing when manually extracting image details,this paper proposes an image contour extraction technology based on U^(2)Net neural network algorithm,and applies it to the digital analysis and design of cultural and creative products of Henan Museum.This method extracts image features from the encoder and uses the decoder to generate segmentation maps based on these features,which greatly saves the time efficiency of manual extraction of cultural relics element modeling,facilitates the subsequent design work of designers.
作者
王嘉玥
夏雅琴
WANG Jiayue;XIA Yaqin
出处
《丝网印刷》
2024年第2期68-70,共3页
Screen Printing
关键词
博物馆文创
神经网络
语义图像分割
文创设计
museum cultural creation
neural network
semantic image segmentation
cultural creation design
