本文深入探讨了多种图像处理技术，包括人脸识别、边缘检测、模板匹配等，并提供了详细的实现方法及代码示例。
/**
* 人脸识别
* @throws IOException
*/
private CascadeClassifier faceDetector;
private void initFaceDetector() throws IOException{
InputStream input = getResources().openRawResource(R.raw.haarcascade_upperbody);
File cascadeDir = this.getDir("cascade", Context.MODE_PRIVATE);
File file = new File(cascadeDir.getAbsoluteFile(),"haarcascade_upperbody.xml");
FileOutputStream output = new FileOutputStream(file);
byte[] buff = new byte[1024];
int len = 0;
while ((len = input.read(buff))!= -1){
output.write(buff,0,len);
}
input.close();
output.close();
faceDetector = new CascadeClassifier(file.getAbsolutePath());
}
/**
* 人脸识别
* @param detector
*/
private void faceDetect(CascadeClassifier detector){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,dst,Imgproc.COLOR_BGRA2GRAY);
MatOfRect faces = new MatOfRect();
detector.detectMultiScale(dst,faces,1.1,15,0,new Size(50,50),new Size());
List<Rect> faceList = faces.toList();
if (faceList.size() >0){
for (Rect rect:faceList){
Imgproc.rectangle(src,rect.tl(),rect.br(),new Scalar(255,0,0,255),2,8,0);
}
}
Utils.matToBitmap(src,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 对象测量
*/
private void measureObjects(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);
int t = 100;
Imgproc.Canny(src,dst,t,t*2,3,false);
List<MatOfPoint> contours = new ArrayList<>();
Mat hierarchy = new Mat();
Imgproc.findContours(dst,contours,hierarchy,Imgproc.RETR_TREE,Imgproc.CHAIN_APPROX_SIMPLE,new Point(0,0));
Imgproc.cvtColor(src,src,Imgproc.COLOR_GRAY2BGR);
double[][] result = new double[contours.size()][2];
for(int i = 0;i <contours.size();i++){
Moments moments = Imgproc.moments(contours.get(i),false);
double m00 = moments.get_m00();
double m10 = moments.get_m10();
double m01 = moments.get_m01();
double x0 = m10 /m00;
double y0 = m01 /m00;
double arclength = Imgproc.arcLength(new MatOfPoint2f(contours.get(i).toArray()),true);
double area = Imgproc.contourArea(contours.get(i));
result[i][0] = arclength;
result[i][1] = area;
Log.d("liubin11 i = ",""+i);
Log.d("liubin11 arclength = ",""+arclength);
Log.d("liubin11 area = ",""+area);
Imgproc.circle(src,new Point(x0,y0),2,new Scalar(255,0,0),2,8,0);
}
Utils.matToBitmap(src,temp);
src.release();
dst.release();
hierarchy.release();
mImageView.setImageBitmap(temp);
}
/**
* 轮廓发现
*/
private void findAndDrawContours(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);
int t = 90;
Imgproc.Canny(src,dst,t,t*2,3,false);
List<MatOfPoint> contours = new ArrayList<MatOfPoint>();
Mat hierarchy = new Mat();
Imgproc.findContours(dst,contours,hierarchy,Imgproc.RETR_TREE,Imgproc.CHAIN_APPROX_SIMPLE,new Point(0,0));
Imgproc.cvtColor(src,src,Imgproc.COLOR_GRAY2BGR);
for (int i =0;i<contours.size();i++){
MatOfPoint points = contours.get(i);
Imgproc.drawContours(src,contours,i,new Scalar(255,0,0),2,8,hierarchy,0,new Point(0,0));
}
Utils.matToBitmap(src,temp);
src.release();
dst.release();
hierarchy.release();
mImageView.setImageBitmap(temp);
}
/**
* 模板匹配
*/
private void templateMat(){
Bitmap tpl = BitmapFactory.decodeResource(this.getResources(),R.drawable.fj);
Bitmap bitmap = BitmapFactory.decodeResource(this.getResources(),R.drawable.fjy);
Mat src = new Mat();
Mat tplMat = new Mat();
Utils.bitmapToMat(bitmap,src);
Utils.bitmapToMat(tpl,tplMat);
int width = bitmap.getWidth() - tpl.getWidth() +1;
int height = bitmap.getHeight() - tpl.getHeight() +1;
Mat result = new Mat(width,height,CvType.CV_32FC1);//result 必须为单通道32位
Imgproc.matchTemplate(src,tplMat,result,Imgproc.TM_CCORR_NORMED);
Core.normalize(result,result,0,1.0,Core.NORM_MINMAX,-1);
Core.MinMaxLocResult minMaxLocResult = Core.minMaxLoc(result);
Point pt = minMaxLocResult.maxLoc;
Imgproc.rectangle(src,pt,new Point(pt.x+tpl.getWidth(),pt.y + tpl.getHeight()),new Scalar(255,0,0,0),2,8,0);
Utils.matToBitmap(src,bitmap);
src.release();
tplMat.release();
result.release();
mImageView.setImageBitmap(bitmap);
}
/**
* 霍夫圆检测
*/
private void houghCircleDet(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);
//dp 分辨率与原来的一样 ；minDist 同心圆直接间隔为5； param1 梯度阈值（基于梯度检测的）；param2 相交点为100次；minRadius 最小半径；maxRadius 最大半径
Imgproc.HoughCircles(src,dst,Imgproc.CV_HOUGH_GRADIENT,1,5,100,50,30,150);
Imgproc.cvtColor(src,src,Imgproc.COLOR_GRAY2BGRA);
double [] circleParams = new double[3];
for (int i =0;i<dst.cols();i++){
circleParams = dst.get(0,i);
Point cp = new Point(circleParams[0],circleParams[1]);
Imgproc.circle(src,cp,(int)circleParams[2],new Scalar(255,0,0,255),2,8,0);
}
Utils.matToBitmap(src,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 霍夫直线检测(从平面坐标到极坐标)
*/
private void houphLinesDet(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Mat lines = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.GaussianBlur(src,src,new Size(3,3),0,0,4);
int t = 20;
Imgproc.Canny(src,dst,t,t*2,3,false);
Mat drawImage = new Mat(src.size(),src.type());
// Imgproc.cvtColor(dst,dst,Imgproc.COLOR_GRAY2BGR);
/*Imgproc.HoughLines(dst,lines,1,Math.PI/180.0,t);
// Imgproc.cvtColor(dst,dst,Imgproc.COLOR_GRAY2BGR);
double[] linep = new double[2];
for (int i =0;i<lines.cols();i++){
linep = lines.get(0,i);
double rho = linep[0];
double theta = linep[1];
double a = Math.cos(theta);
double b = Math.sin(theta);
double x0 = a*rho;
double y0 = b*rho;
Point p1 = new Point(x0+1000*(-b),y0 + 1000*a);
Point p2 = new Point(x0-1000*(-b),y0 - 1000*a);
Imgproc.line(drawImage,p1,p2,new Scalar(255,0,0),2,8,0);
}*/
//直接得到直线
Imgproc.HoughLinesP(dst,lines,1,Math.PI/180.0,t,15,3);
double[] pts = new double[4];
for (int i =0;i<lines.cols();i++){
pts = lines.get(0,i);
Point p1 = new Point(pts[0],pts[1]);
Point p2 = new Point(pts[2],pts[3]);
Imgproc.line(dst,p1,p2,new Scalar(255,0,0),2,8,0);
}
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
lines.release();
mImageView.setImageBitmap(temp);
}
/**
* canny 边缘检测
*
* 实际原理步骤：1、高斯模糊
* 2、梯度计算
* 3、非最大信号压制
* 4、高低阈值链接
* 5、显示
*/
private void cannyEge(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);
Imgproc.GaussianBlur(src,src,new Size(3,3),0,0,4);
int t = 60;
Imgproc.Canny(src,dst,t,t*2,3,false);//threshold 最低阈值，threshold2 最高阈值（一般是最低阈值的2-3倍）
//L2qradient = true 比false识别精度更高
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 计算图像梯度 Sobel 算子 / Scharr 算子
*/
private void sobleGradient(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
int type = 2;
if(type ==1) {//x方向
//ddepth = -1 不变 ； dx = 1 x方向求导 ；
// Imgproc.Sobel(src, dst, -1, 1, 0);//x方向梯度（x方向1阶求导）
// Imgproc.Sobel(src,dst,CvType.CV_16S,1,0);//x方向梯度（x方向求导） 深度为CvType.CV_16S效果比-1要清晰
Imgproc.Scharr(src,dst,CvType.CV_16S,1,0);//Scharr算子，深度设为CvType.CV_16S，若为-1 可能不起作用，因为溢出了
Core.convertScaleAbs(dst,dst);//求绝对值，去除负数
Utils.matToBitmap(dst,temp);
}else if (type ==2) {//y方向
// Imgproc.Sobel(src,dst,-1,0,1);//y方向梯度（y方向求导）
// Imgproc.Sobel(src,dst,CvType.CV_16S,0,1);//y方向梯度（y方向求导） 深度为CvType.CV_16S效果比-1要清晰
Imgproc.Scharr(src,dst,CvType.CV_16S,0,1);//Scharr算子，深度设为CvType.CV_16S，若为-1 可能不起作用，因为溢出了
Core.convertScaleAbs(dst,dst);//求绝对值，去除负数
Utils.matToBitmap(dst,temp);
}else {//x,y 方向梯度结合使用
Mat xdst = new Mat();
Mat ydst = new Mat();
// Imgproc.Sobel(src, xdst, CvType.CV_16S, 1, 0);
// Imgproc.Sobel(src, ydst, CvType.CV_16S, 0, 1);
Imgproc.Scharr(src, xdst, CvType.CV_16S, 1, 0);
Imgproc.Scharr(src, ydst, CvType.CV_16S, 0, 1);
Core.convertScaleAbs(xdst, xdst);
Core.convertScaleAbs(ydst, ydst);
Mat xydst = new Mat();
Core.addWeighted(xdst, 0.5, ydst, 0.5, 30, xydst);//x,y 方向的权重都是0.5
Utils.matToBitmap(xydst, temp);
xdst.release();
ydst.release();
xydst.release();
}
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 直方图均衡化
*/
private void histogramEq(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(selectBitmap,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);//必须转为灰度图，单通道
Imgproc.equalizeHist(src,dst);//直方图均衡化
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(selectBitmap);
}
/**局部阈值：
* 自适应阈值--均值C
* 自适应阈值--高斯C
*/
private void adaptiveThresholdBinary(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);//转为8位灰度单通道图
//自适应阈值
// Imgproc.adaptiveThreshold(src,dst,255,Imgproc.ADAPTIVE_THRESH_MEAN_C,Imgproc.THRESH_BINARY,391,0.0);
//高斯C
Imgproc.adaptiveThreshold(src,dst,255,Imgproc.ADAPTIVE_THRESH_GAUSSIAN_C,Imgproc.THRESH_BINARY,591,0.0);
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**全局阈值：
* 阈值二值化
* 阈值反二值化
* 阈值截断
* 阈值取0
* 阈值取0反
*/
private void thresholdImg(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);//转为8位灰度单通道图
//thresh 为指定阈值，若此值指定则以此值为阈值，若设置为0,最后一个参数加上Imgproc.THRESH_OTSU 则自动计算阈值
// Imgproc.threshold(src,dst,166,255,Imgproc.THRESH_BINARY);//阈值二值化
// Imgproc.threshold(src,dst,166,255,Imgproc.THRESH_BINARY_INV);//阈值反二值化
// Imgproc.threshold(src,dst,0,255,Imgproc.THRESH_BINARY |Imgproc.THRESH_OTSU);//自动阈值二值化
// Imgproc.threshold(src,dst,0,255,Imgproc.THRESH_BINARY_INV |Imgproc.THRESH_OTSU);//自动阈值反二值化
// Imgproc.threshold(src,dst,0,255,Imgproc.THRESH_TRUNC |Imgproc.THRESH_OTSU);//阈值截断
// Imgproc.threshold(src,dst,0,255,Imgproc.THRESH_TOZERO |Imgproc.THRESH_OTSU);//阈值取0
Imgproc.threshold(src,dst,0,255,Imgproc.THRESH_TOZERO_INV |Imgproc.THRESH_OTSU);//阈值取0反
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 直线识别(横线)
*/
private void morphLineDetection(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);//转成灰度图（单通道）
Imgproc.threshold(src,src,0,255,Imgproc.THRESH_BINARY_INV | Imgproc.THRESH_OTSU);//二值化处理
//strElement 结构元素形态的选取，此处Imgproc.MORPH_RECT 为特殊矩形，高为1
Mat strElement = Imgproc.getStructuringElement(Imgproc.MORPH_RECT,new Size(20,1),new Point(-1,-1));
Imgproc.morphologyEx(src,dst,Imgproc.MORPH_OPEN,strElement);//开操作
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 开（先腐蚀再膨胀），闭（先膨胀再腐蚀） 操作
*/
private void openOrClose(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2GRAY);//转成灰度图（单通道）
Imgproc.threshold(src,src,0,255,Imgproc.THRESH_BINARY_INV | Imgproc.THRESH_OTSU);//二值化处理
//strElement 结构元素形态的选取
Mat strElement = Imgproc.getStructuringElement(Imgproc.MORPH_RECT,new Size(3,3),new Point(-1,-1));
// Imgproc.morphologyEx(src,dst,Imgproc.MORPH_OPEN,strElement);//开操作
Imgproc.morphologyEx(src,dst,Imgproc.MORPH_CLOSE,strElement);//闭操作
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 图片腐蚀(既最小值滤波)和膨胀（既最大值滤波）
*/
private void erodeOrDilate(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
//结构形态的选取
Mat strElement = Imgproc.getStructuringElement(Imgproc.MORPH_RECT,new Size(3,3),new Point(-1,-1));
// Imgproc.erode(src,dst,strElement,new Point(-1,-1),5);//腐蚀， 最后一个参数，表示执行几次腐蚀
Imgproc.dilate(src,dst,strElement,new Point(-1,-1),5);//膨胀，最后一个值表示执行1次膨胀
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 自定义滤波器
*/
private void customFilter(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Mat kernel = getCustomOperator();
Imgproc.filter2D(src,dst,-1,kernel,new Point(-1,-1),0.0,4);
Utils.matToBitmap(dst,temp);
kernel.release();
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
private Mat getCustomOperator(){
Mat kernel = new Mat(3,3,CvType.CV_32FC1);//自定义滤波算子，必须是基数 如此处的（3，3）
kernel.put(0,0,1.0/9.0,1.0/9.0,1.0/9.0,1.0/9.0,1.0/9.0,1.0/9.0,1.0/9.0,1.0/9.0,1.0/9.0);//模糊
// kernel.put(0,0,-1,-1,-1,-1,8,-1,-1,-1,-1);//拉普拉斯边缘
// kernel.put(0,0,-1,-1,-1,-1,9,-1,-1,-1,-1);//拉普拉斯边缘锐化
return kernel;
}
/**
* 双边模糊
*/
private void biBlur(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,src,Imgproc.COLOR_BGRA2BGR);//4通道转3通道,双边模糊只支持3通道，或单通道，而Bitmap是4通道的，要做转换
Imgproc.bilateralFilter(src,dst,15,150,15,4);//双边模糊
Mat kernel = new Mat(3,3,CvType.CV_16S);
kernel.put(0,0,0,-1,0,-1,5,-1,0,-1,0);
Imgproc.filter2D(dst,dst,-1,kernel,new Point(-1,-1),0,4);//锐化处理
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 高斯模糊 速度没有均值模糊快
*/
private void gaussianBlur(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
//Size 中若不为0则以size中的计算，若为0则已sigmaX,Y来计算模糊范围
Imgproc.GaussianBlur(src,dst,new Size(3,3),0,0,4);
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 均值模糊
*/
private void meanBlur(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Mat dst = new Mat();
Utils.bitmapToMat(temp,src);
//Size 表示内核的大小（也叫算子） 都必须为正数和奇数，模糊范围如（9，9） 表示x方向9个像素，y方向9个像素，如（1，9） 表示只在y方向做模糊
//Point （-1，-1）表示默认为内核中心点
//最后一个参数 用于推断图像外部像素的某种边界模式。注意它有默认值BORDER_DEFAULT
Imgproc.blur(src,dst,new Size(1,35),new Point(-1,-1),4);
Utils.matToBitmap(dst,temp);
src.release();
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 截取图片指定区域
*/
private void getROIArea(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Rect roi = new Rect(200,150,200,300);
Bitmap roimap = Bitmap.createBitmap(roi.width,roi.height,Bitmap.Config.ARGB_8888);
Mat src = new Mat();
Utils.bitmapToMat(temp,src);
Mat roiMat = src.submat(roi);//截取指定位置图片
Mat roiDstMat = new Mat();
Imgproc.cvtColor(roiMat,roiDstMat,Imgproc.COLOR_BGR2GRAY);//转灰度图
Utils.matToBitmap(roiDstMat,roimap);
roiDstMat.release();
roiMat.release();
src.release();
mImageView.setImageBitmap(roimap);
}
/**
* 创建空白的Mat
*/
private void creatMat(){
Bitmap bitmap = Bitmap.createBitmap(400,600,Bitmap.Config.ARGB_8888);
// Mat dst = new Mat(bitmap.getHeight(),bitmap.getWidth(),CvType.CV_8UC1,new Scalar(100));//CU_8UC1 表示一个通道，所以后面的Scalar也只有一个通道
Mat dst = new Mat(bitmap.getHeight(),bitmap.getWidth(),CvType.CV_8UC4,new Scalar(255,0,0,255));//通道顺序agbr
Utils.matToBitmap(dst,bitmap);
dst.release();
mImageView.setImageBitmap(bitmap);
}
/**
* 转灰度图
*/
private void convert3Gray(){
Mat src = new Mat();
Mat dst = new Mat();
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Utils.bitmapToMat(temp,src);
Imgproc.cvtColor(src,dst,Imgproc.COLOR_BGR2GRAY);
Utils.matToBitmap(dst,temp);
src.release();//必须释放内存
dst.release();
mImageView.setImageBitmap(temp);
}
/**
* 利用Mat像素反转
*/
private void invert(){
Mat src = new Mat();
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Utils.bitmapToMat(temp,src);
//pixel options
int width = src.cols();
int height = src.rows();
int cnum = src.channels();//通道数，如 ARGB_8888 占4个通道
byte[] bgra = new byte[cnum];
long startTime = System.currentTimeMillis();
for (int row = 0; row <height;row++){
for (int col = 0;col < width;col++){
src.get(row,col,bgra);//将row行，col列的像素，保存到bgra中
for (int i = 0;i< cnum;i++){
bgra[i] = (byte)(255 -bgra[i]&0xff);//0xff 使得byte 不发生有符号拓展，使用一个掩码来限制. 主要对负数的限制
}
src.put(row,col,bgra);
}
}
long end = System.currentTimeMillis() - startTime;
Log.d(TAG,"inver total time "+ end);
Utils.matToBitmap(src,temp);
src.release();
mImageView.setImageBitmap(temp);
}
/**
* 利用Mat像素反转
*/
private void invert2(){
Mat src = new Mat();
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Utils.bitmapToMat(temp,src);
long startTime = System.currentTimeMillis();
Core.bitwise_not(src,src);
long end = System.currentTimeMillis() - startTime;
Log.d(TAG,"invert2 total time "+ end);
Utils.matToBitmap(src,temp);
src.release();
mImageView.setImageBitmap(temp);
}
/**
* 像素减法
*/
private void subStract(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Utils.bitmapToMat(temp,src);
Mat whiteImg = new Mat(src.size(),src.type(), Scalar.all(255));//创建一个白色的Mat 大小和src必须一样
long startTime = System.currentTimeMillis();
Core.subtract(whiteImg,src,src);
long end = System.currentTimeMillis() - startTime;
Log.d(TAG,"subStract end = "+end);
Utils.matToBitmap(src,temp);
src.release();
whiteImg.release();
mImageView.setImageBitmap(temp);
}
/**
* 像素加法
*/
private void add(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Utils.bitmapToMat(temp,src);
Mat whiteImg = new Mat(src.size(),src.type(), Scalar.all(255));//创建一个白色的Mat 大小和src必须一样
long startTime = System.currentTimeMillis();
// Core.add(whiteImg,src,src);
Core.addWeighted(whiteImg,0.5,src,0.5,0,src);
long end = System.currentTimeMillis() - startTime;
Log.d(TAG,"subStract end = "+end);
Utils.matToBitmap(src,temp);
src.release();
whiteImg.release();
mImageView.setImageBitmap(temp);
}
/**
* 调整对比度和亮度
*/
private void adjustContrast(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
Mat src = new Mat();
Utils.bitmapToMat(temp,src);
src.convertTo(src, CvType.CV_32F);//转成32位的浮点数，为后面能使用浮点数计算所要做的步骤
Mat whiteImg = new Mat(src.size(),src.type(), Scalar.all(1.25));//创建一个白色的Mat 大小和src必须一样
Mat bwImg = new Mat(src.size(),src.type(),Scalar.all(30));
long startTime = System.currentTimeMillis();
Core.multiply(whiteImg,src,src);//对各像素先乘以1.25
Core.add(bwImg,src,src);//再对各个像素加个30
long end = System.currentTimeMillis() - startTime;
Log.d(TAG,"subStract end = "+end);
src.convertTo(src,CvType.CV_8U);//将src转为CV_8U，才能转为Bitmap
Utils.matToBitmap(src,temp);
src.release();
whiteImg.release();
bwImg.release();
mImageView.setImageBitmap(temp);
}
/**
* 利用Bitmap 的缓存中做像素反转
*/
private void bitmapInvert(){
Bitmap temp = selectBitmap.copy(selectBitmap.getConfig(),true);
int width = temp.getWidth();
int height = temp.getHeight();
int [] pixels = new int[width*height];
temp.getPixels(pixels,0,width,0,0,width,height);
int index = 0;
int a = 0,r = 0,g = 0,b = 0;
long startTime = System.currentTimeMillis();
for (int row = 0;row < height;row ++){
index = row*width;
for (int col = 0;col < width ;col ++){
int pixel = pixels[index];
a = (pixel>>24)&0xff;
r = (pixel>>16)&0xff;
g = (pixel>>8)&0xff;
b = (pixel&0xff);
r = 255 -r;//反转
g = 255 -g;
b = 255 -b;
pixel = ((a&0xff)<<24) | ((r&0xff)<<16) | ((g&0xff)<<8) | (b&0xff);
pixels[index] = pixel;
index++;
}
}
long end = System.currentTimeMillis() - startTime;
Log.d(TAG,"bitmap getPixels end = "+ end);
temp.setPixels(pixels,0,width,0,0,width,height);
mImageView.setImageBitmap(temp);
}
Open CV 3.4基础API使用
最新推荐文章于 2025-05-19 08:53:21 发布
扫一扫
4648
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
