Example Binary Image

Binary Image Processing

Binary images are images where each pixel can take on two values, typically represented by 0 or 1. Binary images are easy to compute and fast to process, which makes them popular in many applications.

Example File: BinaryImageExample.java

Concepts:

• Image Thresholding
• Morphological Operations
• Binary Labeling
• Pixel Math
• Image Rendering

Relevant Applets:

• Binary Operations
• Binary Segmentation

Basic Example

In this example a threshold is computed for the input image dynamically and the resulting binary image shown.

Output of binary image example.

	public static void binaryExample( BufferedImage image )
	{
		// convert into a usable format
		ImageFloat32 input = ConvertBufferedImage.convertFrom(image,null,ImageFloat32.class);
		ImageUInt8 binary = new ImageUInt8(input.width,input.height);

		// the mean pixel value is often a reasonable threshold when creating a binary image
		float mean = PixelMath.sum(input)/(input.width*input.height);

		// create a binary image
		ThresholdImageOps.threshold(input,binary,mean,true);

		// Render the binary image for output and display it in a window
		BufferedImage visualBinary = VisualizeBinaryData.renderBinary(binary,null);
		ShowImages.showWindow(visualBinary,"Binary Image");
	}

Labeled Example

Here clustered of blobs are detected and arbitrarily assigned labels.

Output of binary label example.

	public static void labeledExample( BufferedImage image )
	{
		// convert into a usable format
		ImageFloat32 input = ConvertBufferedImage.convertFrom(image,null,ImageFloat32.class);
		ImageUInt8 binary = new ImageUInt8(input.width,input.height);
		ImageSInt32 blobs = new ImageSInt32(input.width,input.height);

		// the mean pixel value is often a reasonable threshold when creating a binary image
		float mean = PixelMath.sum(input)/(input.width*input.height);

		// create a binary image
		ThresholdImageOps.threshold(input,binary,mean,true);

		// remove small blobs through erosion and dilation
		// The null in the input indicates that it should internally declare the work image it needs
		// this is less efficient, but easier to code.
		binary = BinaryImageOps.erode8(binary,null);
		binary = BinaryImageOps.dilate8(binary, null);

		// Detect blobs inside the binary image and assign labels to them
		int numBlobs = BinaryImageOps.labelBlobs4(binary,blobs);

		// Render the binary image for output and display it in a window
		BufferedImage visualized = VisualizeBinaryData.renderLabeled(blobs, numBlobs, null);
		ShowImages.showWindow(visualized,"Labeled Image");
	}