org.neuroph.imgrec

Class ImageSampler

java.lang.Object

org.neuroph.imgrec.ImageSampler

public class ImageSampler
extends java.lang.Object

This class uses a given Robot object to sample images from the screen at an arbitrary sampling resolution. The coordinates of the scanning rectangle is (x, y) expressed as a fraction of the screen resolution, with the width also being a fraction. For example, when the screen is 800x600 resolution, to scan a rectangle with xy coordinates (400,300) to xy (600, 400), use the following rectangle as an argument: new Rectangle2D.Double(0.5, 0.5, 0.25, 0.16667) Scanning using sampling is faster than scanning using a screenshot, but is prone to introduce tearing and shearing into the scanned image. Scanning from a screenshot has no tearing or shearing at the cost of speed. Also provides method for downsampling (scaling) image. This class is based on the code from tileclassification by Jon Tait.

Constructor Summary

Constructors

Constructor and Description
ImageSampler()

Method Summary

Modifier and Type	Method and Description
static Image	downSampleImage(Dimension samplingResolution, Image image)
static Image	downSampleImage(Dimension samplingResolution, Image bigImage, int returnImageType)

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

ImageSampler

public ImageSampler()

Method Detail

downSampleImage

public static Image downSampleImage(Dimension samplingResolution,
                                    Image bigImage,
                                    int returnImageType)

downSampleImage

public static Image downSampleImage(Dimension samplingResolution,
                                    Image image)