拆分图像目录Split Image Directory

本主题介绍如何使用 Azure 机器学习设计器中的“拆分图像目录”模块将图像目录的图像划分为两个不同的集。This topic describes how to use the Split Image Directory module in Azure Machine Learning designer, to divide the images of an image directory into two distinct sets.

当需要将图像数据划分为训练集和测试集时,此模块特别有用。This module is particularly useful when you need to separate image data into training and testing sets.

如何配置“拆分图像目录”模块How to configure Split Image Directory

  1. 将“拆分图像目录”模块添加到管道。Add the Split Image Directory module to your pipeline. 可以在“计算机视觉/图像数据转换”类别下找到此模块。You can find this module under 'Computer Vision/Image Data Transformation' category.

  2. 将它连接到输出为其图像目录的模块。Connect it to module of which the output is image directory.

  3. 输入 第一个输出中的图像的比例 ,以指定要放入左分割区的数据百分比(默认为 0.9)。Input Fraction of images in the first output to specify the percentage of data to put in the left split, by default 0.9.

技术说明Technical notes

预期输入Expected inputs

名称Name 类型Type 说明Description
输入图像目录Input image directory ImageDirectoryImageDirectory 要拆分的图像目录Image directory to split

模块参数Module parameters

名称Name 类型Type 范围Range 可选Optional 说明Description 默认Default
第一个输出中的图像比例Fraction of images in the first output FloatFloat 0-10-1 必须Required 第一个输出中的图像比例Fraction of images in the first output 0.90.9

OutputsOutputs

名称Name 类型Type 说明Description
输出图像目录 1Output image directory1 ImageDirectoryImageDirectory 包含所选图像的图像目录Image directory that contains selected images
输出图像目录 2Output image directory2 ImageDirectoryImageDirectory 包含所有其他图像的图像目录Image directory that contains all other images

后续步骤Next steps

请参阅 Azure 机器学习的可用模块集See the set of modules available to Azure Machine Learning.