图像数据标签标识系统设计研究
Research on the Design of Image Data Label Identification System
杨明福 1徐亮 1杨政 1王亮 1官永杨1
作者信息
- 1. 贵州江南航天信息网络通信有限公司,贵州遵义 563000
- 折叠
摘要
图像数据是机器视觉、机器学习等人工智能领域的重要数据源,而图像数据标签则是开展图像数据应用的重要一环,尤其伴随着深度学习等前沿技术兴起,图像数据的标签完整性及准确性对于图像数据应用起到至关重要的作用.以图像数据高效处理的需求为牵引,结合图像标签标识流程及方法,开展图像数据标签标识系统模块化设计与搭建工作,并将图像标识功能单元进行封装处理,同时开展系统交互接口,并实现图像数据标识功能单元的快速组合与应用,经过测试试验,较人工标注和其他方法而言,设计的系统具有效率高、标注方式齐全等优势.
Abstract
Image data is an important data source in artificial intelligence fields such as machine vision and machine learning,and image data labels are an important part of image data applications.Especially with the rise of cutting-edge technologies such as deep learning,the integrity and accuracy of image data labels play a crucial role in image data ap-plications.This project is driven by the need for efficient processing of image data,combined with the process and methods of image label identification,to carry out modular design and construction of the image data label identification system,and encapsulate the image identification functional units,carry out system interaction interfaces,and achieve rapid combination and application of image data identification functional units.After testing and testing,compared to manual annotation and other methods,the system designed in this paper has advantages such as high efficiency,complete annotation methods.
关键词
数据标签/图像数据/模块化设计Key words
data labels/image data/modular design引用本文复制引用
基金项目
贵州省科技支撑计划(黔科合支撑[2023]一般323)
遵义市科技创新人才团队培养计划(遵市科人才[2022]6号)
出版年
2024
NETL
NSTL
万方数据