PPTX DOCX XLSX PDF ODP
Aspose.Imaging Java
DICOM

通過 Java 從 DICOM 中刪除背景

構建您自己的 Java 應用程序以使用服務器端 API 從 DICOM 文件中刪除背景。

 

Aspose.Imaging Java Aspose.Imaging Java

Overview

如何使用 Java 刪除 DICOM 文件中的背景

從圖像中去除背景涉及隔離前景對象，這是一項需要對象識別的任務。存在多種方法來識別 DICOM 格式照片中的對象。對於具有統一顏色背景的簡單圖像，自動方法就足夠了。然而，對於具有多個或部分合併人物的照片，預先標記對象就變得必要。這涉及指定要刪除的矩形區域和對像類型。在復雜的情況下，Cloud API 可以實現自動對象檢測。 Cloud API 提供了一個雲應用程序，能夠識別照片中的對象，利用生成的輪廓來去除背景。去除後，可以平滑圖形留下的邊緣以提高圖像質量。為了刪除 DICOM 文件中的背景，我們將使用 Aspose.Imaging for Java API 是一個功能豐富、功能強大 易於使用的 Java 平台圖像處理和轉換 API。您可以直接從 Maven 並通過將以下配置添加到 pom.xml 將其安裝在基於 Maven 的項目中。

Repository

<repository>
<id>AsposeJavaAPI</id>
<name>Aspose Java API</name>
<url>https://repository.aspose.com/repo/</url>
</repository>

Dependency

<dependency>
<groupId>com.aspose</groupId>
<artifactId>aspose-imaging</artifactId>
<version>version of aspose-imaging API</version>
<classifier>jdk16</classifier>
</dependency>

通過 Java 從 DICOM 中刪除背景的步驟

你需要 aspose-imaging-version-jdk16.jar 在您自己的環境中嘗試以下工作流程。

  • 使用 Image.load 方法加載 DICOM 文件
  • 移除背景；
  • 以 Aspose.Imaging 支持的格式將圖像保存到光盤

系統要求

所有主要操作系統都支持 Java 的 Aspose.Imaging。只需確保您具有以下先決條件。

  • 已安裝 JDK 1.6 或更高版本。
 

刪除 DICOM 圖像中的背景 - Java

import com.aspose.imaging.*;
import com.aspose.imaging.fileformats.png.PngColorType;
import com.aspose.imaging.imageoptions.PngOptions;
import com.aspose.imaging.masking.IMaskingSession;
import com.aspose.imaging.masking.ImageMasking;
import com.aspose.imaging.masking.options.*;
import com.aspose.imaging.masking.result.MaskingResult;
import com.aspose.imaging.sources.FileCreateSource;
import java.io.File;
import java.util.Arrays;
import java.util.LinkedList;
import java.util.List;
import java.util.function.Consumer;
//Most common example to demonstrate background change/remove tool
removeBackgroundGenericExample();
// Folder that contains images to process
static final String templatesFolder = "c:\\Data\\";
public static void removeBackgroundProcessingWithManualRectangles()
{
List<String> rasterFormats = Arrays.asList("jpg", "png", "bmp", "apng",
"dicom", "jp2", "j2k", "tga", "webp", "tif", "gif");
List<String> vectorFormats = Arrays.asList("svg", "otg", "odg", "eps",
"wmf", "emf", "wmz", "emz", "cmx", "cdr");
List<String> allFormats = new LinkedList<>(rasterFormats);
allFormats.addAll(vectorFormats);
allFormats.forEach(new Consumer<String>()
{
@Override
public void accept(String formatExt)
{
String inputFile = templatesFolder + "couple." + formatExt;
boolean isVectorFormat = vectorFormats.contains(formatExt);
//Need to rasterize vector formats before background remove
if (isVectorFormat)
{
inputFile = rasterizeVectorImage(formatExt, inputFile);
}
String outputFile = templatesFolder + "remove_background_manual_rectangles." + formatExt;
System.out.println("Processing " + formatExt);
try (RasterImage image = (RasterImage) Image.load(inputFile))
{
//Additional code examples can be found at
//https://docs.aspose.com/imaging/java/remove-background-from-images/#graph-cut-auto-masking-using-imagingcloud-api
AutoMaskingGraphCutOptions maskingOptions = new AutoMaskingGraphCutOptions();
maskingOptions.setFeatheringRadius(2);
maskingOptions.setMethod(SegmentationMethod.GraphCut);
AutoMaskingArgs maskingArgs = new AutoMaskingArgs();
maskingArgs.setObjectsRectangles(new Rectangle[]
{
// girl's bound box
new Rectangle(87, 47, 123, 308),
// boy's bound box
new Rectangle(180, 24, 126, 224)
});
maskingOptions.setArgs(maskingArgs);
PngOptions pngOptions = new PngOptions();
pngOptions.setColorType(PngColorType.TruecolorWithAlpha);
pngOptions.setSource(new FileCreateSource(outputFile, false));
maskingOptions.setExportOptions(pngOptions);
IMaskingSession maskingSession = new ImageMasking(image)
.createSession(maskingOptions);
try
{
// first run of segmentation
maskingSession.decompose().dispose();
AutoMaskingArgs argsWithUserMarkers = new AutoMaskingArgs();
argsWithUserMarkers.setObjectsPoints(new Point[][]
{
// background markers
null,
// foreground markers
new UserMarker()
// boy's head
.addPoint(218, 48, 10)
// girl's head
.addPoint(399, 66, 10)
// girs's body
.addPoint(158, 141, 10)
.addPoint(158, 209, 20)
.addPoint(115, 225, 5)
.getPoints()
});
try (MaskingResult maskingResult = maskingSession
.improveDecomposition(argsWithUserMarkers))
{
try (Image resultImage = maskingResult.get_Item(1).getImage())
{
resultImage.save();
}
}
}
finally
{
maskingSession.dispose();
}
}
//Remove rasterized vector image
if (isVectorFormat)
{
new File(inputFile).delete();
}
}
});
}
public static void removeBackgroundAutoProcessingWithAssumedObjects()
{
List<String> rasterFormats = Arrays.asList("jpg", "png", "bmp", "apng",
"dicom", "jp2", "j2k", "tga", "webp", "tif", "gif");
List<String> vectorFormats = Arrays.asList("svg", "otg", "odg", "eps",
"wmf", "emf", "wmz", "emz", "cmx", "cdr");
List<String> allFormats = new LinkedList<>(rasterFormats);
allFormats.addAll(vectorFormats);
allFormats.forEach(new Consumer<String>() {
@Override
public void accept(String formatExt)
{
String inputFile = templatesFolder + "couple." + formatExt;
boolean isVectorFormat = vectorFormats.contains(formatExt);
//Need to rasterize vector formats before background remove
if (isVectorFormat)
{
inputFile = rasterizeVectorImage(formatExt, inputFile);
}
String outputFile = templatesFolder
+ "remove_background_auto_assumed_objects." + formatExt;
System.out.println("Processing " + formatExt);
try (RasterImage image = (RasterImage) Image.load(inputFile))
{
//Additional code examples can be found at
//https://docs.aspose.com/imaging/java/remove-background-from-images/#graph-cut-auto-masking-using-imagingcloud-api
AutoMaskingGraphCutOptions maskingOptions = new AutoMaskingGraphCutOptions();
final LinkedList<AssumedObjectData> assumedObjects = new LinkedList<>();
// girl's bound box
assumedObjects.add(
new AssumedObjectData(DetectedObjectType.Human,
new Rectangle(87, 47, 123, 308)));
// boy's bound box
assumedObjects.add(
new AssumedObjectData(DetectedObjectType.Human,
new Rectangle(180, 24, 126, 224)));
maskingOptions.setAssumedObjects(assumedObjects);
maskingOptions.setCalculateDefaultStrokes(true);
maskingOptions.setFeatheringRadius(1);
maskingOptions.setMethod(SegmentationMethod.GraphCut);
PngOptions pngOptions = new PngOptions();
pngOptions.setColorType(PngColorType.TruecolorWithAlpha);
pngOptions.setSource(new FileCreateSource(outputFile, false));
maskingOptions.setExportOptions(pngOptions);
maskingOptions.setBackgroundReplacementColor(Color.getGreen());
try (MaskingResult maskingResult = new ImageMasking(image)
.decompose(maskingOptions))
{
try (Image resultImage = maskingResult.get_Item(1).getImage())
{
resultImage.save();
}
}
}
//Remove rasterized vector image
if (isVectorFormat)
{
new File(inputFile).delete();
}
}
});
}
public static void removeBackgroundAutoProcessing()
{
List<String> rasterFormats = Arrays.asList("jpg", "png", "bmp", "apng",
"dicom", "jp2", "j2k", "tga", "webp", "tif", "gif");
List<String> vectorFormats = Arrays.asList("svg", "otg", "odg", "eps",
"wmf", "emf", "wmz", "emz", "cmx", "cdr");
List<String> allFormats = new LinkedList<>(rasterFormats);
allFormats.addAll(vectorFormats);
allFormats.forEach(new Consumer<String>() {
@Override
public void accept(String formatExt)
{
String inputFile = templatesFolder + "couple." + formatExt;
boolean isVectorFormat = vectorFormats.contains(formatExt);
//Need to rasterize vector formats before background remove
if (isVectorFormat)
{
inputFile = rasterizeVectorImage(formatExt, inputFile);
}
String outputFile = templatesFolder + "remove_background_auto." + formatExt;
System.out.println("Processing " + formatExt);
try (RasterImage image = (RasterImage) Image.load(inputFile))
{
//Additional code examples can be found at
//https://docs.aspose.com/imaging/java/remove-background-from-images/#graph-cut-auto-masking-using-imagingcloud-api
AutoMaskingGraphCutOptions maskingOptions = new AutoMaskingGraphCutOptions();
maskingOptions.setFeatheringRadius(1);
maskingOptions.setMethod(SegmentationMethod.GraphCut);
PngOptions pngOptions = new PngOptions();
pngOptions.setColorType(PngColorType.TruecolorWithAlpha);
pngOptions.setSource(new FileCreateSource(outputFile, false));
maskingOptions.setExportOptions(pngOptions);
maskingOptions.setBackgroundReplacementColor(Color.getGreen());
try (MaskingResult maskingResult = new ImageMasking(image)
.decompose(maskingOptions))
{
try (Image resultImage = maskingResult.get_Item(1).getImage())
{
resultImage.save();
}
}
}
//Remove rasterized vector image
if (isVectorFormat)
{
new File(inputFile).delete();
}
}
});
}
public static void removeBackgroundGenericExample()
{
List<String> rasterFormats = Arrays.asList("jpg", "png", "bmp", "apng",
"dicom", "jp2", "j2k", "tga", "webp", "tif", "gif");
List<String> vectorFormats = Arrays.asList("svg", "otg", "odg", "eps",
"wmf", "emf", "wmz", "emz", "cmx", "cdr");
List<String> allFormats = new LinkedList<>(rasterFormats);
allFormats.addAll(vectorFormats);
allFormats.forEach(new Consumer<String>() {
@Override
public void accept(String formatExt)
{
String inputFile = templatesFolder + "couple." + formatExt;
boolean isVectorFormat = vectorFormats.contains(formatExt);
//Need to rasterize vector formats before background remove
if (isVectorFormat)
{
inputFile = rasterizeVectorImage(formatExt, inputFile);
}
String outputFile = templatesFolder + "remove_background." + formatExt;
System.out.println("Processing " + formatExt);
try (RasterImage image = (RasterImage) Image.load(inputFile))
{
//Additional code examples can be found at
//https://docs.aspose.com/imaging/java/remove-background-from-images/#graph-cut-auto-masking-using-imagingcloud-api
AutoMaskingGraphCutOptions maskingOptions = new AutoMaskingGraphCutOptions();
maskingOptions.setCalculateDefaultStrokes(true);
maskingOptions.setFeatheringRadius(1);
maskingOptions.setMethod(SegmentationMethod.GraphCut);
PngOptions pngOptions = new PngOptions();
pngOptions.setColorType(PngColorType.TruecolorWithAlpha);
pngOptions.setSource(new FileCreateSource(outputFile, false));
maskingOptions.setExportOptions(pngOptions);
maskingOptions.setBackgroundReplacementColor(Color.getGreen());
try (MaskingResult maskingResult = new ImageMasking(image)
.decompose(maskingOptions))
{
try (Image resultImage = maskingResult.get_Item(1).getImage())
{
resultImage.save();
}
}
}
//Remove rasterized vector image
if (isVectorFormat)
{
new File(inputFile).delete();
}
}
});
}
private static String rasterizeVectorImage(String formatExt, String inputFile)
{
String outputFile = templatesFolder + "rasterized."+ formatExt + ".png";
try (Image image = Image.load(inputFile))
{
image.save(outputFile, new PngOptions());
}
return outputFile;
}
class UserMarker
{
private final List<Point> list = new LinkedList<>();
public UserMarker addPoint(int left, int top, int radius)
{
for (int y = top - radius; y <= top + radius; y++)
{
for (int x = left - radius; x <= left + radius; x++)
{
this.list.add(new Point(x, y));
}
}
return this;
}
public Point[] getPoints()
{
return this.list.toArray(new Point[0]);
}
}
view raw remove-change-background-generic-examples.java hosted with ❤ by GitHub
 

    • 關於 Java API 的 Aspose.Imaging

    Aspose.Imaging API 是一種圖像處理解決方案，用於在應用程序中創建、修改、繪製或轉換圖像（照片）。它提供：跨平台的圖像處理，包括但不限於各種圖像格式之間的轉換（包括統一的多頁或多幀圖像處理）、繪圖等修改、使用圖形基元、轉換（調整大小、裁剪、翻轉和旋轉） 、二值化、灰度、調整）、高級圖像處理功能（過濾、抖動、遮罩、去偏斜）和內存優化策略。它是一個獨立的庫，不依賴任何軟件進行圖像操作。可以在項目中使用原生 API 輕鬆添加高性能圖像轉換功能。這些是 100% 私有的本地 API，圖像在您的服務器上處理。

    通過在線應用刪除 DICOM 中的背景

    通過訪問我們的 Live Demos 網站 移除 DICOM 文檔中的背景。 現場演示有以下好處

      無需下載或設置任何東西
      無需編寫任何代碼
      只需上傳您的 DICOM 文件並點擊“立即刪除背景”按鈕
      立即獲取生成文件的下載鏈接

    DICOM 什麼是 DICOM 文件格式

    DICOM 是 Digital Imaging and Communications in Medicine 的首字母縮寫詞，屬於醫學信息學領域。 DICOM 是文件格式定義和網絡通信協議的結合。 DICOM 使用 .DCM 擴展名。 .DCM 以兩種不同的格式存在，即格式 1.x 和格式 2.x。 DCM Format 1.x 還提供兩個普通版本和擴展版本。 DICOM 用於集成來自不同供應商的打印機、服務器、掃描儀等醫療成像設備，還包含每個患者的唯一識別數據。如果 DICOM 文件能夠接收 DICOM 格式的圖像數據，則它們可以在兩方之間共享。 DICOM的通信部分是應用層協議，實體之間使用TCP/IP進行通信。 HTTP 和 HTTPS 協議用於 DICOM 的 Web 服務。 Web 服務支持的版本是 1.0、1.1、2 或更高版本。

    閱讀更多

    其他支持的刪除背景格式

    使用 Java，可以輕鬆地從不同格式中刪除背景，包括。

    APNG (動畫便攜式網絡圖形)
    BMP (位圖圖片)
    ICO (窗口圖標)
    JPG (聯合攝影專家組)
    DIB (設備無關位圖)
    DJVU (圖形格式)
    DNG (數碼相機圖像)
    EMF (增強的元文件格式)
    EMZ (Windows 壓縮增強元文件)
    GIF (圖形交換格式)
    JP2 (JPEG 2000)
    J2K (小波壓縮圖像)
    PNG (便攜式網絡圖形)
    TIFF (標記圖像格式)
    WEBP (光柵網絡圖像)
    WMF (微軟視窗元文件)
    WMZ (壓縮的 Windows Media Player 皮膚)
    TGA (塔加圖形)
    SVG (可縮放矢量圖形)
    EPS (封裝的 PostScript 語言)
    CDR (矢量繪圖圖像)
    CMX (Corel 交換圖像)
    OTG (開放文檔標準)
    ODG (Apache OpenOffice 繪圖格式)
    AVIF