|
// Licensed to the .NET Foundation under one or more agreements.
// The .NET Foundation licenses this file to you under the MIT license.
// See the LICENSE file in the project root for more information.
using System;
using System.Buffers;
using System.Collections.Generic;
using System.Diagnostics.Contracts;
using System.Drawing;
using System.IO;
using System.Linq;
using System.Text;
using Microsoft.ML;
using Microsoft.ML.CommandLine;
using Microsoft.ML.Data;
using Microsoft.ML.Internal.Utilities;
using Microsoft.ML.Runtime;
using Microsoft.ML.Transforms;
using Microsoft.ML.Transforms.Image;
[assembly: LoadableClass(ImageLoadingTransformer.Summary, typeof(IDataTransform), typeof(ImageLoadingTransformer), typeof(ImageLoadingTransformer.Options), typeof(SignatureDataTransform),
ImageLoadingTransformer.UserName, "ImageLoaderTransform", "ImageLoader")]
[assembly: LoadableClass(ImageLoadingTransformer.Summary, typeof(IDataTransform), typeof(ImageLoadingTransformer), null, typeof(SignatureLoadDataTransform),
ImageLoadingTransformer.UserName, ImageLoadingTransformer.LoaderSignature)]
[assembly: LoadableClass(typeof(ImageLoadingTransformer), null, typeof(SignatureLoadModel), "", ImageLoadingTransformer.LoaderSignature)]
[assembly: LoadableClass(typeof(IRowMapper), typeof(ImageLoadingTransformer), null, typeof(SignatureLoadRowMapper), "", ImageLoadingTransformer.LoaderSignature)]
namespace Microsoft.ML.Data
{
/// <summary>
/// <see cref="ITransformer"/> resulting from fitting a <see cref="ImageLoadingEstimator"/>.
/// </summary>
public sealed class ImageLoadingTransformer : OneToOneTransformerBase
{
internal sealed class Column : OneToOneColumn
{
internal static Column Parse(string str)
{
Contracts.AssertNonEmpty(str);
var res = new Column();
if (res.TryParse(str))
return res;
return null;
}
internal bool TryUnparse(StringBuilder sb)
{
Contracts.AssertValue(sb);
return TryUnparseCore(sb);
}
}
internal sealed class Options : TransformInputBase
{
[Argument(ArgumentType.Multiple | ArgumentType.Required, HelpText = "New column definition(s) (optional form: name:src)", Name = "Column", ShortName = "col", SortOrder = 1)]
public Column[] Columns;
[Argument(ArgumentType.AtMostOnce, HelpText = "Folder where to search for images", ShortName = "folder")]
public string ImageFolder;
}
internal const string Summary = "Load images from files.";
internal const string UserName = "Image Loader Transform";
internal const string LoaderSignature = "ImageLoaderTransform";
/// <summary>
/// The folder to load the images from.
/// </summary>
public readonly string ImageFolder;
/// <summary>
/// The flag for DataViewType for the image. If Type true, it is a VectorDataView of bytes else it is an ImageDataView type.
/// If no options are specified, it defaults to false for ImageDataView type.
/// </summary>
private readonly bool _useImageType;
/// <summary>
/// The columns passed to this <see cref="ITransformer"/>.
/// </summary>
internal IReadOnlyCollection<(string outputColumnName, string inputColumnName)> Columns => ColumnPairs.AsReadOnly();
/// <summary>
/// Initializes a new instance of <see cref="ImageLoadingTransformer"/>.
/// </summary>
/// <param name="env">The host environment.</param>
/// <param name="imageFolder">Folder where to look for images.</param>
/// <param name="columns">Names of input and output columns.</param>
internal ImageLoadingTransformer(IHostEnvironment env, string imageFolder = null, params (string outputColumnName, string inputColumnName)[] columns)
: this(env, imageFolder, type: true, columns)
{
}
/// <summary>
/// Initializes a new instance of <see cref="ImageLoadingTransformer"/>.
/// </summary>
/// <param name="env">The host environment.</param>
/// <param name="imageFolder">Folder where to look for images.</param>
/// <param name="type">Image type flag - true for ImageDataViewType or false for VectorDataViewType. Defaults to true i.e. ImageDataViewType if not specified.</param>
/// <param name="columns">Names of input and output columns.</param>
internal ImageLoadingTransformer(IHostEnvironment env, string imageFolder = null, bool type = true, params (string outputColumnName, string inputColumnName)[] columns)
: base(Contracts.CheckRef(env, nameof(env)).Register(nameof(ImageLoadingTransformer)), columns)
{
// Throws ArgumentException if given imageFolder path is invalid. Note: imageFolder may be null or empty in this case.
if (String.IsNullOrEmpty(imageFolder))
ImageFolder = null;
else
{
if (Directory.Exists(imageFolder))
ImageFolder = Path.GetFullPath(imageFolder);
else
throw new ArgumentException(String.Format("Directory \"{0}\" does not exist.", imageFolder));
}
_useImageType = type;
}
// Factory method for SignatureDataTransform.
internal static IDataTransform Create(IHostEnvironment env, Options options, IDataView data)
{
return new ImageLoadingTransformer(env, options.ImageFolder, options.Columns.Select(x => (x.Name, x.Source ?? x.Name)).ToArray())
.MakeDataTransform(data);
}
// Factory method for SignatureLoadModel.
private static ImageLoadingTransformer Create(IHostEnvironment env, ModelLoadContext ctx)
{
Contracts.CheckValue(env, nameof(env));
env.CheckValue(ctx, nameof(ctx));
ctx.CheckAtModel(GetVersionInfo());
return new ImageLoadingTransformer(env.Register(nameof(ImageLoadingTransformer)), ctx);
}
private ImageLoadingTransformer(IHost host, ModelLoadContext ctx)
: base(host, ctx)
{
// *** Binary format ***
// <base>
// int: id of image folder
ImageFolder = ctx.LoadStringOrNull();
if (ctx.Header.ModelVerWritten >= 0x00010003) // do a version check
_useImageType = ctx.Reader.ReadBoolean();
else
_useImageType = true; // It is an ImageDataViewType
}
// Factory method for SignatureLoadDataTransform.
private static IDataTransform Create(IHostEnvironment env, ModelLoadContext ctx, IDataView input)
=> Create(env, ctx).MakeDataTransform(input);
// Factory method for SignatureLoadRowMapper.
private static IRowMapper Create(IHostEnvironment env, ModelLoadContext ctx, DataViewSchema inputSchema)
=> Create(env, ctx).MakeRowMapper(inputSchema);
private protected override void CheckInputColumn(DataViewSchema inputSchema, int col, int srcCol)
{
if (!(inputSchema[srcCol].Type is TextDataViewType))
throw Host.ExceptSchemaMismatch(nameof(inputSchema), "input", ColumnPairs[col].inputColumnName, TextDataViewType.Instance.ToString(), inputSchema[srcCol].Type.ToString());
}
private protected override void SaveModel(ModelSaveContext ctx)
{
Host.CheckValue(ctx, nameof(ctx));
ctx.CheckAtModel();
ctx.SetVersionInfo(GetVersionInfo());
// *** Binary format ***
// <base>
// int: id of image folder
base.SaveColumns(ctx);
ctx.SaveStringOrNull(ImageFolder);
ctx.Writer.Write(_useImageType);
}
private static VersionInfo GetVersionInfo()
{
return new VersionInfo(
modelSignature: "IMGLOADR",
//verWrittenCur: 0x00010001, // Initial
verWrittenCur: 0x00010003, // Added support for output type as byte array
verReadableCur: 0x00010003,
verWeCanReadBack: 0x00010003,
loaderSignature: LoaderSignature,
loaderAssemblyName: typeof(ImageLoadingTransformer).Assembly.FullName);
}
private protected override IRowMapper MakeRowMapper(DataViewSchema schema) => new Mapper(this, schema, _useImageType);
private sealed class Mapper : OneToOneMapperBase
{
private readonly ImageLoadingTransformer _parent;
private readonly bool _type;
public Mapper(ImageLoadingTransformer parent, DataViewSchema inputSchema, bool type)
: base(parent.Host.Register(nameof(Mapper)), parent, inputSchema)
{
_type = type;
_parent = parent;
}
protected override Delegate MakeGetter(DataViewRow input, int iinfo, Func<int, bool> activeOutput, out Action disposer)
{
disposer = null;
// Check for the type of Image, if true load images as ImageDataViewType else load images as VBuffer<bytes>
if (_type)
return MakeGetterImageDataViewType(input, iinfo, activeOutput, out disposer);
else
return MakeGetterVectorDataViewByteType(input, iinfo, activeOutput, out disposer);
}
private Delegate MakeGetterImageDataViewType(DataViewRow input, int iinfo, Func<int, bool> activeOutput, out Action disposer)
{
Contracts.AssertValue(input);
Contracts.Assert(0 <= iinfo && iinfo < _parent.ColumnPairs.Length);
var lastImage = default(MLImage);
disposer = () =>
{
if (lastImage != null)
{
lastImage.Dispose();
lastImage = null;
}
};
var getSrc = input.GetGetter<ReadOnlyMemory<char>>(input.Schema[ColMapNewToOld[iinfo]]);
ReadOnlyMemory<char> src = default;
ValueGetter<MLImage> del =
(ref MLImage dst) =>
{
if (dst != null)
{
dst.Dispose();
dst = null;
}
getSrc(ref src);
if (src.Length > 0)
{
string path = src.ToString();
if (!string.IsNullOrWhiteSpace(_parent.ImageFolder))
path = Path.Combine(_parent.ImageFolder, path);
// to avoid locking file, use the construct below to load the image
var bytes = File.ReadAllBytes(path);
var ms = new MemoryStream(bytes);
dst = MLImage.CreateFromStream(ms);
dst.Tag = path;
}
lastImage = dst;
};
return del;
}
private Delegate MakeGetterVectorDataViewByteType(DataViewRow input, int iinfo, Func<int, bool> activeOutput, out Action disposer)
{
Contracts.AssertValue(input);
Contracts.Assert(0 <= iinfo && iinfo < _parent.ColumnPairs.Length);
disposer = null;
var getSrc = input.GetGetter<ReadOnlyMemory<char>>(input.Schema[ColMapNewToOld[iinfo]]);
ReadOnlyMemory<char> src = default;
ValueGetter<VBuffer<byte>> del =
(ref VBuffer<byte> dst) =>
{
getSrc(ref src);
if (src.Length > 0)
{
string path = src.ToString();
if (!string.IsNullOrWhiteSpace(_parent.ImageFolder))
path = Path.Combine(_parent.ImageFolder, path);
if (!TryLoadDataIntoBuffer(path, ref dst))
{
var editor = VBufferEditor.Create(ref dst, 0); //Empty Image
dst = editor.Commit();
}
}
else
{
var editor = VBufferEditor.Create(ref dst, 0);
dst = editor.Commit();
}
};
return del;
}
private static bool TryLoadDataIntoBuffer(string path, ref VBuffer<byte> imgData)
{
int count = -1;
int bytesread = -1;
// bufferSize == 1 used to avoid unnecessary buffer in FileStream
using (FileStream fs = new FileStream(path, FileMode.Open, FileAccess.Read, FileShare.Read, bufferSize: 1))
{
long fileLength = fs.Length;
if (fileLength > int.MaxValue)
throw new IOException($"File {path} too big to open.");
else if (fileLength == 0)
{
byte[] imageBuffer;
// Some file systems (e.g. procfs on Linux) return 0 for length even when there's content.
// Thus we need to assume 0 doesn't mean empty.
imageBuffer = File.ReadAllBytes(path);
count = imageBuffer.Length;
imgData = new VBuffer<byte>(count, imageBuffer);
return (count > 0);
}
count = (int)fileLength;
var editor = VBufferEditor.Create(ref imgData, count);
bytesread = ReadToEnd(fs, editor.Values);
imgData = editor.Commit();
return (count > 0);
}
}
private static int ReadToEnd(System.IO.Stream stream, Span<byte> bufferSpan)
{
int totalBytesRead = 0;
int bytesRead;
int chunksize = 4096; // Most optimal size for buffer, friendly to CPU's L1 cache
var bufferPool = ArrayPool<byte>.Shared;
byte[] readBuffer = bufferPool.Rent(chunksize);
try
{
var readBufferSpan = readBuffer.AsSpan();
var srcSpan = readBufferSpan;
while ((bytesRead = stream.Read(readBuffer, 0, readBuffer.Length)) > 0)
{
if (bytesRead != srcSpan.Length)
srcSpan = readBufferSpan.Slice(0, bytesRead);
var dstSpan = bufferSpan.Slice(totalBytesRead, bytesRead);
Contract.Assert(srcSpan.Length == dstSpan.Length);
srcSpan.CopyTo(dstSpan);
totalBytesRead += bytesRead;
}
}
finally
{
// don't use the reference to the buffer after returning it!
bufferPool.Return(readBuffer);
}
return totalBytesRead;
}
private DataViewType GetDataViewType()
{
if (_type)
return new ImageDataViewType();
else
return new VectorDataViewType(NumberDataViewType.Byte);
}
protected override DataViewSchema.DetachedColumn[] GetOutputColumnsCore()
=> _parent.ColumnPairs.Select(x => new DataViewSchema.DetachedColumn(x.outputColumnName, GetDataViewType(), null)).ToArray();
}
}
/// <summary>
/// <see cref="IEstimator{TTransformer}"/> for the <see cref="ImageLoadingTransformer"/>.
/// </summary>
/// <remarks>
/// <format type="text/markdown">< |
/// | Output column data type | <xref:Microsoft.ML.Data.MLImage> |
/// | Required NuGet in addition to Microsoft.ML | Microsoft.ML.ImageAnalytics |
/// | Exportable to ONNX | No |
///
/// The resulting <xref:Microsoft.ML.Data.ImageLoadingTransformer> creates a new column, named as specified in the output column name parameters, and
/// loads in it images specified in the input column.
/// Loading is the first step of almost every pipeline that does image processing, and further analysis on images.
/// The images to load need to be in the formats supported by <xref:Microsoft.ML.Data.MLImage> implementation.
/// For end-to-end image processing pipelines, and scenarios in your applications, see the
/// [examples](https://github.com/dotnet/machinelearning-samples/tree/main/samples/csharp/getting-started) in the machinelearning-samples github repository.</a>
///
/// Check the See Also section for links to usage examples.
/// ]]>
/// </format>
/// </remarks>
/// <seealso cref="ImageEstimatorsCatalog.LoadImages(TransformsCatalog, string, string, string)" />
public sealed class ImageLoadingEstimator : TrivialEstimator<ImageLoadingTransformer>
{
private readonly bool _type;
/// <summary>
/// Load images in memory.
/// </summary>
/// <param name="env">The host environment.</param>
/// <param name="imageFolder">Folder where to look for images.</param>
/// <param name="columns">Names of input and output columns.</param>
internal ImageLoadingEstimator(IHostEnvironment env, string imageFolder, params (string outputColumnName, string inputColumnName)[] columns)
: this(env, new ImageLoadingTransformer(env, imageFolder, true, columns), true)
{
}
/// <summary>
/// Load images in memory.
/// </summary>
/// <param name="env">The host environment.</param>
/// <param name="imageFolder">Folder where to look for images.</param>
/// <param name="type">Image type flag - true for ImageDataViewType or false for VectorDataView. Defaults to true i.e ImageDataViewType if not specified.</param>
/// <param name="columns">Names of input and output columns.</param>
internal ImageLoadingEstimator(IHostEnvironment env, string imageFolder, bool type = true, params (string outputColumnName, string inputColumnName)[] columns)
: this(env, new ImageLoadingTransformer(env, imageFolder, type, columns), type)
{
}
internal ImageLoadingEstimator(IHostEnvironment env, ImageLoadingTransformer transformer, bool type = true)
: base(Contracts.CheckRef(env, nameof(env)).Register(nameof(ImageLoadingEstimator)), transformer)
{
_type = type;
}
/// <summary>
/// Returns the <see cref="SchemaShape"/> of the schema which will be produced by the transformer.
/// Used for schema propagation and verification in a pipeline.
/// </summary>
public override SchemaShape GetOutputSchema(SchemaShape inputSchema)
{
Host.CheckValue(inputSchema, nameof(inputSchema));
var result = inputSchema.ToDictionary(x => x.Name);
foreach (var (outputColumnName, inputColumnName) in Transformer.Columns)
{
if (!inputSchema.TryFindColumn(inputColumnName, out var col))
throw Host.ExceptSchemaMismatch(nameof(inputSchema), "input", inputColumnName);
if (!(col.ItemType is TextDataViewType) || col.Kind != SchemaShape.Column.VectorKind.Scalar)
throw Host.ExceptSchemaMismatch(nameof(inputSchema), "input", inputColumnName, TextDataViewType.Instance.ToString(), col.GetTypeString());
if (_type)
result[outputColumnName] = new SchemaShape.Column(outputColumnName, SchemaShape.Column.VectorKind.Scalar, new ImageDataViewType(), false);
else
result[outputColumnName] = new SchemaShape.Column(outputColumnName, SchemaShape.Column.VectorKind.VariableVector, NumberDataViewType.Byte, false);
}
return new SchemaShape(result.Values);
}
}
}
|