PPTX DOCX XLSX PDF ODP
Aspose.Imaging  Python
WMF

使用 Python 将 WMF 图像卡通化

创建 Python 应用程序以通过服务器 API 卡通化 WMF 图像和照片

如何使用 Python 将 WMF 图像和照片卡通化

我们会自动对卡通图像做出反应,因为它们能够唤起怀旧感。在平面设计领域,卡通风格的图像是营销文章中常见的关键元素。这种卡通化效果涉及将照片肖像转换为手绘效果、调整亮度、转换为黑白、使用调色板以及合并各种编辑技术来制作复杂的视觉效果。一套图像过滤器,包括“AdjustBrightness”、“BinarizeFixed”、“Filter”、“ReplaceColor”和“ApplyMask”,使用户能够实现这些转换。这些过滤器可用于已下载的原始格式图像和照片。卡通风格的图像适用于各种网页的插图目的,为科学文章注入活力,并使内容对用户更具吸引力,从而增加网站的流量。要使用 WMF 图像生成卡通效果,我们将采用 Aspose.Imaging for Python via .NET API 是一个功能丰富、功能强大且易于使用的图像处理和转换 API,适用于 Python 平台。您可以使用系统命令中的以下命令安装它。

系统命令行

>> pip install aspose-imaging-python-net

通过 Python 对 WMF 进行卡通化的步骤

您需要 aspose-imaging-python-net 在您自己的环境中尝试以下工作流程。

  • 使用 Image.Load 方法加载 WMF 文件 +卡通化图像;
  • 以 Aspose.Imaging 支持的格式将压缩图像保存到光盘

系统要求

所有主要操作系统都支持 Python 的 Aspose.Imaging。只需确保您具有以下先决条件。

  • 带有 .NET Core 运行时的 Microsoft Windows / Linux。
  • Python 和 PyPi 包管理器。
 

Cartoonify WMF 图像 - Python

using Aspose.Imaging;
using Aspose.Imaging.FileFormats.Png;
using Aspose.Imaging.ImageFilters.FilterOptions;
using Aspose.Imaging.ImageOptions;
using Aspose.Imaging.Masking;
using Aspose.Imaging.Masking.Options;
using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
string templatesFolder = @"c:\Users\USER\Downloads";
Cartoonify();
void Cartoonify()
{
FilterImages(image =>
{
using (var processedImage = new PngImage(image))
{
image.Resize(image.Width * 2, image.Height, ResizeType.LeftTopToLeftTop);
processedImage.Cartoonify();
var gr = new Graphics(image);
gr.DrawImage(processedImage, processedImage.Width, 0);
gr.DrawLine(new Pen(Color.DarkRed, 3), processedImage.Width, 0, processedImage.Width, image.Height);
}
}, "cartoonify");
}
string RasterizeVectorImage(string formatExt, string inputFile)
{
string outputFile = Path.Combine(templatesFolder, $"rasterized.{formatExt}.png");
using (var image = Image.Load(inputFile))
{
image.Save(outputFile, new PngOptions());
}
return outputFile;
}
void FilterImages(Action<RasterImage> doFilter, string filterName)
{
List<string> rasterFormats = new List<string>() { "jpg", "png", "bmp", "apng", "dicom",
"jp2", "j2k", "tga", "webp", "tif", "gif", "ico" };
List<string> vectorFormats = new List<string>() { "svg", "otg", "odg", "eps", "wmf", "emf", "wmz", "emz", "cmx", "cdr" };
List<string> allFormats = new List<string>(rasterFormats);
allFormats.AddRange(vectorFormats);
allFormats.ForEach(
formatExt =>
{
var inputFile = Path.Combine(templatesFolder, $"template.{formatExt}");
bool isVectorFormat = vectorFormats.IndexOf(formatExt) > -1;
//Need to rasterize vector formats before background remove
if (isVectorFormat)
{
inputFile = RasterizeVectorImage(formatExt, inputFile);
}
var outputFile = Path.Combine(templatesFolder, $"{filterName}_{formatExt}.png");
Console.WriteLine($"Processing {formatExt}");
using (var image = (RasterImage)Image.Load(inputFile))
{
doFilter(image);
//If image is multipage save each page to png to demonstrate results
if (image is IMultipageImage multiPage && multiPage.PageCount > 1)
{
for (var pageIndex = 0; pageIndex < multiPage.PageCount; pageIndex++)
{
string fileName = $"{filterName}_page{pageIndex}_{formatExt}.png";
multiPage.Pages[pageIndex].Save(templatesFolder + fileName, new PngOptions());
File.Delete(templatesFolder + fileName);
}
}
else
{
image.Save(outputFile, new PngOptions());
File.Delete(outputFile);
}
}
//Remove rasterized vector image
if (isVectorFormat)
{
File.Delete(inputFile);
}
}
);
}
static class ImageFilterExtensions
{
public static void Cartoonify(this RasterImage image)
{
using var outlines = image.DetectOutlines(Color.Black);
image.AdjustBrightness(30);
image.Filter(image.Bounds, new MedianFilterOptions(7));
var gr = new Graphics(image);
gr.DrawImage(outlines, Point.Empty);
}
public static RasterImage DetectOutlines(this RasterImage image, Color outlineColor)
{
var outlines = new PngImage(image);
outlines
.GetDataContext()
.ApplyConvolutionFilter(ConvolutionFilterOptions.Blur)
.ApplyConvolutionFilter(ConvolutionFilterOptions.Outline)
.ApplyData();
outlines.BinarizeFixed(30);
ImageMasking.ApplyMask(outlines, outlines, new MaskingOptions() { BackgroundReplacementColor = Color.Transparent });
outlines.ReplaceColor(Color.FromArgb(255, 255, 255), 0, outlineColor);
outlines.ApplyConvolutionFilter(ConvolutionFilterOptions.Blur);
return outlines;
}
public static RasterImage ApplyOperationToRasterImage(this RasterImage image, Action<RasterImage> operation)
{
if (image is IMultipageImage multipage)
{
foreach (var page in multipage.Pages)
{
operation.Invoke((RasterImage)page);
}
}
else
{
operation.Invoke(image);
}
return image;
}
public static RasterImage ApplyFilter(this RasterImage image, FilterOptionsBase filterOptions)
{
return image.ApplyOperationToRasterImage(img =>
{
img.Filter(img.Bounds, filterOptions);
});
}
public static RasterImage ApplyConvolutionFilter(this RasterImage image, ConvolutionFilterOptions filterOptions)
{
return image.ApplyOperationToRasterImage(img =>
{
var pixelsLoader = new ImagePixelsLoader(img.Bounds);
img.LoadPartialArgb32Pixels(img.Bounds, pixelsLoader);
var outBuffer = new PixelBuffer(img.Bounds, new int[img.Width * img.Height]);
ConvolutionFilter.DoFiltering(pixelsLoader.PixelsBuffer, outBuffer, filterOptions);
img.SaveArgb32Pixels(outBuffer.Rectangle, outBuffer.Pixels);
});
}
public static IImageDataContext GetDataContext(this RasterImage image)
{
IPixelBuffer GetImageBuffer(RasterImage img)
{
var pixelsLoader = new ImagePixelsLoader(img.Bounds);
img.LoadPartialArgb32Pixels(img.Bounds, pixelsLoader);
return pixelsLoader.PixelsBuffer;
}
if (image is IMultipageImage multipage)
{
return new MultipageDataContext(
multipage.Pages.Select(page => new ImageDataContext((RasterImage)page)
{
Buffer = GetImageBuffer((RasterImage)page)
}));
}
return new ImageDataContext(image)
{
Buffer = GetImageBuffer(image)
};
}
public static IImageDataContext ApplyToDataContext(this IImageDataContext dataContext,
Func<IPixelBuffer, IPixelBuffer> processor)
{
if (dataContext is MultipageDataContext multipage)
{
foreach (var context in multipage)
{
context.Buffer = processor.Invoke(context.Buffer);
}
}
if (dataContext is ImageDataContext imageDataContext)
{
imageDataContext.Buffer = processor.Invoke(imageDataContext.Buffer);
}
return dataContext;
}
public static IImageDataContext ApplyConvolutionFilter(this IImageDataContext dataContext,
ConvolutionFilterOptions filterOptions)
{
return dataContext.ApplyToDataContext(buffer =>
{
var outBuffer = new PixelBuffer(buffer.Rectangle, new int[buffer.Rectangle.Width * buffer.Rectangle.Height]);
ConvolutionFilter.DoFiltering(buffer, outBuffer, filterOptions);
return outBuffer;
});
}
}
class ConvolutionFilter
{
public static void DoFiltering(
IPixelBuffer inputBuffer,
IPixelBuffer outputBuffer,
ConvolutionFilterOptions options)
{
var factor = options.Factor;
var bias = options.Bias;
var kernel = options.Kernel;
var filterWidth = kernel.GetLength(1);
var filterCenter = (filterWidth - 1) / 2;
int x, y;
int filterX, filterY, filterPx, filterPy, filterYPos, pixel;
double r, g, b, kernelValue;
int top = inputBuffer.Rectangle.Top;
int bottom = inputBuffer.Rectangle.Bottom;
int left = inputBuffer.Rectangle.Left;
int right = inputBuffer.Rectangle.Right;
for (y = top; y < bottom; y++)
{
for (x = left; x < right; x++)
{
r = 0;
g = 0;
b = 0;
for (filterY = -filterCenter; filterY <= filterCenter; filterY++)
{
filterYPos = filterY + filterCenter;
filterPy = filterY + y;
if (filterPy >= top && filterPy < bottom)
{
for (filterX = -filterCenter; filterX <= filterCenter; filterX++)
{
filterPx = filterX + x;
if (filterPx >= left && filterPx < right)
{
kernelValue = kernel[filterYPos, filterX + filterCenter];
pixel = inputBuffer[filterPx, filterPy];
r += ((pixel >> 16) & 0xFF) * kernelValue;
g += ((pixel >> 8) & 0xFF) * kernelValue;
b += (pixel & 0xFF) * kernelValue;
}
}
}
}
r = (factor * r) + bias;
g = (factor * g) + bias;
b = (factor * b) + bias;
r = r > 255 ? 255 : (r < 0 ? 0 : r);
g = g > 255 ? 255 : (g < 0 ? 0 : g);
b = b > 255 ? 255 : (b < 0 ? 0 : b);
outputBuffer[x, y] = ((inputBuffer[x, y] >> 24) << 24) | ((byte)r << 16) | ((byte)g << 8) | (byte)b;
}
}
}
}
class ConvolutionFilterOptions
{
public double Factor { get; set; } = 1.0;
public int Bias { get; set; } = 0;
public double[,] Kernel { get; set; }
public static ConvolutionFilterOptions Blur
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { 1, 2, 1 }, { 2, 4, 2 }, { 1, 2, 1 } },
Factor = 0.25 * 0.25
};
}
}
public static ConvolutionFilterOptions Sharpen
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { 0, -1, 0 }, { -1, 5, -1 }, { 0, -1, 0 } }
};
}
}
public static ConvolutionFilterOptions Emboss
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { -2, -1, 0 }, { -1, 1, 1 }, { 0, 1, 2 } }
};
}
}
public static ConvolutionFilterOptions Outline
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { -1, -1, -1 }, { -1, 8, -1 }, { -1, -1, -1 } }
};
}
}
public static ConvolutionFilterOptions BottomSobel
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { -1, -2, -1 }, { 0, 0, 0 }, { 1, 2, 1 } }
};
}
}
public static ConvolutionFilterOptions TopSobel
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { 1, 2, 1 }, { 0, 0, 0 }, { -1, -2, -1 } }
};
}
}
public static ConvolutionFilterOptions LeftSobel
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { 1, 0, -1 }, { 2, 0, -2 }, { 1, 0, -1 } }
};
}
}
public static ConvolutionFilterOptions RightSobel
{
get
{
return new ConvolutionFilterOptions
{
Kernel = new double[,] { { -1, 0, 1 }, { -2, 0, 2 }, { -1, 0, 1 } }
};
}
}
}
interface IImageDataContext
{
void ApplyData();
}
class ImageDataContext : IImageDataContext
{
public ImageDataContext(RasterImage image)
{
this.Image = image;
}
public RasterImage Image { get; }
public IPixelBuffer Buffer { get; set; }
public void ApplyData()
{
this.Buffer.SaveToImage(this.Image);
}
}
class MultipageDataContext : List<ImageDataContext>, IImageDataContext
{
public MultipageDataContext(IEnumerable<ImageDataContext> enumerable) : base(enumerable)
{
}
public void ApplyData()
{
foreach (var context in this)
{
context.ApplyData();
}
}
}
class ImagePixelsLoader : IPartialArgb32PixelLoader
{
public ImagePixelsLoader(Aspose.Imaging.Rectangle rectangle)
{
this.PixelsBuffer = new CompositePixelBuffer(rectangle);
}
public CompositePixelBuffer PixelsBuffer { get; }
public void Process(Aspose.Imaging.Rectangle pixelsRectangle, int[] pixels, Point start, Point end)
{
this.PixelsBuffer.AddPixels(pixelsRectangle, pixels);
}
}
interface IPixelBuffer
{
Aspose.Imaging.Rectangle Rectangle { get; }
int this[int x, int y]
{
get;
set;
}
void SaveToImage(RasterImage image);
}
class PixelBuffer : IPixelBuffer
{
public PixelBuffer(Aspose.Imaging.Rectangle rectangle, int[] pixels)
{
this.Rectangle = rectangle;
this.Pixels = pixels;
}
public Aspose.Imaging.Rectangle Rectangle { get; }
public int[] Pixels { get; }
public int this[int x, int y]
{
get => this.Pixels[this.GetIndex(x, y)];
set => this.Pixels[this.GetIndex(x, y)] = value;
}
public void SaveToImage(RasterImage image)
{
image.SaveArgb32Pixels(this.Rectangle, this.Pixels);
}
public bool Contains(int x, int y)
{
return this.Rectangle.Contains(x, y);
}
private int GetIndex(int x, int y)
{
x -= this.Rectangle.Left;
y -= this.Rectangle.Top;
return x + y * this.Rectangle.Width;
}
}
class CompositePixelBuffer : IPixelBuffer
{
private readonly List<PixelBuffer> _buffers = new List<PixelBuffer>();
public CompositePixelBuffer(Aspose.Imaging.Rectangle rectangle)
{
this.Rectangle = rectangle;
}
public Aspose.Imaging.Rectangle Rectangle { get; }
public int this[int x, int y]
{
get => this.GetBuffer(x, y)[x, y];
set => this.GetBuffer(x, y)[x, y] = value;
}
public void SaveToImage(RasterImage image)
{
foreach (var pixelBuffer in this._buffers)
{
pixelBuffer.SaveToImage(image);
}
}
public IEnumerable<PixelBuffer> Buffers => this._buffers;
public void AddPixels(Aspose.Imaging.Rectangle rectangle, int[] pixels)
{
if (this.Rectangle.IntersectsWith(rectangle))
{
this._buffers.Add(new PixelBuffer(rectangle, pixels));
}
}
private PixelBuffer GetBuffer(int x, int y)
{
return this._buffers.First(b => b.Contains(x, y));
}
}
 
  • 关于 Python API 的 Aspose.Imaging

    Aspose.Imaging API 是一种图像处理解决方案,用于在应用程序中创建、修改、绘制或转换图像(照片)。它提供:跨平台的图像处理,包括但不限于各种图像格式之间的转换(包括统一的多页或多帧图像处理)、绘图等修改、使用图形基元、转换(调整大小、裁剪、翻转和旋转) 、二值化、灰度、调整)、高级图像处理功能(过滤、抖动、遮罩、去偏斜)和内存优化策略。它是一个独立的库,不依赖任何软件进行图像操作。可以在项目中使用原生 API 轻松添加高性能图像转换功能。这些是 100% 私有的本地 API,图像在您的服务器上处理。

    通过在线应用程序卡通化 WMF

    通过访问我们的 Live Demos 网站 对 WMF 文档进行卡通化。 现场演示有以下好处

      无需下载或设置任何东西
      无需编写任何代码
      只需上传您的 WMF 文件并点击 "Cartoonify now" 按钮
      立即获取生成文件的下载链接

    WMF 什么是 WMF 文件格式

    带有 WMF 扩展名的文件代表 Microsoft Windows 元文件 (WMF),用于存储矢量以及位图格式的图像数据。更准确地说,WMF 属于与设备无关的图形文件格式的矢量文件格式类别。 Windows 图形设备接口 (GDI) 使用存储在 WMF 文件中的函数在屏幕上显示图像。后来发布了 WMF 的更增强版本,称为增强元文件 (EMF),使该格式的功能更加丰富。实际上,WMF 类似于 SVG。

    阅读更多

    其他支持的卡通化格式

    使用 Python,可以轻松卡通化不同的格式,包括。

    APNG (动画便携式网络图形)
    BMP (位图图片)
    ICO (窗口图标)
    JPG (联合摄影专家组)
    JPEG (联合摄影专家组)
    DIB (设备无关位图)
    DICOM (数码影像与通讯)
    DJVU (图形格式)
    DNG (数码相机图像)
    EMF (增强的元文件格式)
    EMZ (Windows 压缩增强元文件)
    GIF (图形交换格式)
    JP2 (JPEG 2000)
    J2K (小波压缩图像)
    PNG (便携式网络图形)
    TIFF (标记图像格式)
    TIF (标记图像格式)
    WEBP (光栅 Web 图像)
    WMZ (压缩的 Windows Media Player 皮肤)
    TGA (塔加图形)
    SVG (可缩放矢量图形)
    EPS (封装的 PostScript 语言)
    CDR (矢量绘图图像)
    CMX (Corel 交换图像)
    OTG (开放文档标准)
    ODG (Apache OpenOffice 绘图格式)