前言:
从C++ 复现到Java上,真心无语!
单单是搜索来的就不够详尽,更别说Java了。
本次复现案例来自
https://blog.csdn.net/u011503970/article/details/18623913/
环境及版本:JDK 13.0.1 OpenCV 4.1.2 Windows7 x64
下面是Java上关于重点的DFT了。看看说明!
void org.opencv.core.Core.dft(Mat src, Mat dst)
Performs a forward or inverse Discrete Fourier transform of a 1D or 2D floating-point array. The function cv::dft performs one of the following:。。。。。。
Parameters:
src input array that could be real or complex.
dst output array whose size and type depends on the flags . nonzeroRows rows of the input array (#DFT_INVERSE is not set) or only the first nonzeroRows of the output array (#DFT_INVERSE is set) contain non-zeros, thus, the function can handle the rest of the rows more efficiently and save some time; this technique is very useful for calculating array cross-correlation or convolution using DFT. SEE: dct , getOptimalDFTSize , mulSpectrums, filter2D , matchTemplate , flip , cartToPolar , magnitude , phase
上面的说明能看出啥,折腾半天才知道传入的是单通道 的灰度图
所以你得先将原图转到灰度图
Imgproc.cvtColor(img, gray, Imgproc.COLOR_RGB2GRAY);
后面的就好办了,直接上完整代码:
import java.util.ArrayList;
import java.util.List;import org.opencv.core.Core;
import org.opencv.core.CvType;
import org.opencv.core.Mat;
import org.opencv.core.Rect;
import org.opencv.core.Scalar;
import org.opencv.highgui.HighGui;
import org.opencv.imgcodecs.Imgcodecs;
import org.opencv.imgproc.Imgproc;/**
* @Description: DFT图像傅立叶变换 复现
* @author: Administrator
* @date: 2019年11月14日 下午12:49:19
*/
public class TestDft {public Mat dftStart(Mat img) {
img.convertTo(img, CvType.CV_32FC1);
System.out.println("img类型: " + img.type() + " " + img.channels());
int M = Core.getOptimalDFTSize(img.rows()); // 获得最佳DFT尺寸,为2的次方
int N = Core.getOptimalDFTSize(img.cols()); // 同上
Mat padded = new Mat();
System.out.println("padded 类型: " + padded.size() + " " + padded.type() + " " + padded.channels());
// Mat padded1 = new Mat();
Core.copyMakeBorder(img, padded, 0, M - img.rows(), 0, N - img.cols(), Core.BORDER_CONSTANT, new Scalar(0)); // opencv中的边界扩展函数,提供多种方式扩展
System.out.println("padded 类型: " + padded.size() + " " + padded.type() + " " + padded.channels());
// Imgcodecs.imwrite("d:\\123\\table_test7.jpg", padded);
// padded.convertTo(padded1, CvType.CV_32FC1);
System.out.println("padded 类型: " + padded.size() + " " + padded.type() + " " + padded.channels());
// System.out.println("padded1 类型: " + padded1.type() + " " + padded1.channels());
List<Mat> planes = new ArrayList<Mat>(); // Mat 数组,第一个为扩展后的图像,一个为空图像,
planes.add(padded);
planes.add(Mat.zeros(padded.size(), CvType.CV_32FC1));
Mat complexImg = new Mat();
// Mat.zeros(padded.size(), CvType.CV_32F);System.out
.println("complexImg 类型: " + complexImg.size() + " " + complexImg.type() + " " + complexImg.channels());
Core.merge(planes, complexImg); // 合并成一个Mat
System.out
.println("complexImg 类型: " + complexImg.size() + " " + complexImg.type() + " " + complexImg.channels());Core.dft(complexImg, complexImg); // FFT变换, dft需要一个2通道的Mat
// compute log(1 + sqrt(Re(DFT(img))**2 + Im(DFT(img))**2))
Core.split(complexImg, planes); // 分离通道, planes[0] 为实数部分,planes[1]为虚数部分
Core.magnitude(planes.get(0), planes.get(1), planes.get(0)); // 求模
Mat mag = planes.get(0);
Core.add(mag, new Scalar(1), mag);
// mag += new Scalar(1);
Core.log(mag, mag); // 模的对数// crop the spectrum, if it has an odd number of rows or columns
mag = new Mat(mag, new Rect(0, 0, mag.cols() & -2, mag.rows() & -2)); // 保证偶数的边长int cx = mag.cols() / 2;
int cy = mag.rows() / 2;// rearrange the quadrants of Fourier image //对傅立叶变换的图像进行重排,4个区块,从左到右,从上到下
// 分别为q0, q1, q2, q3
// so that the origin is at the image center // 对调q0和q3, q1和q2
Mat tmp = new Mat();
Mat q0 = new Mat(mag, new Rect(0, 0, cx, cy));
Mat q1 = new Mat(mag, new Rect(cx, 0, cx, cy));
Mat q2 = new Mat(mag, new Rect(0, cy, cx, cy));
Mat q3 = new Mat(mag, new Rect(cx, cy, cx, cy));q0.copyTo(tmp);
q3.copyTo(q0);
tmp.copyTo(q3);q1.copyTo(tmp);
q2.copyTo(q1);
tmp.copyTo(q2);// Core.normalize(mag, mag, 0, 1, Core.NORM_MINMAX); // 规范化值到 0~1 显示图片的需要 归一化
System.out.println("mag 类型: " + mag.size() + " " + mag.type() + " " + mag.channels());
mag.convertTo(mag, CvType.CV_8U);
return mag;}
public static void main(String[] args) {
System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
Mat img = Imgcodecs.imread("d:\\123\\20191113173328.jpg");
img = Imgcodecs.imread("d:\\123\\20191106112608.jpg");
// img = Imgcodecs.imread("d:\\123\\20191106152434.jpg");
Mat gray = new Mat();
Imgproc.cvtColor(img, gray, Imgproc.COLOR_RGB2GRAY);
TestDft t = new TestDft();
Mat dst = t.dftStart(gray);
// Imgcodecs.imwrite("d:\\123\\test1.jpg", dst);
HighGui.imshow("原图", img);
HighGui.imshow("dft效果图", dst);
HighGui.waitKey(0);
HighGui.destroyAllWindows();
System.exit(0);
}}
效果图:
待完善。。。