OpenCV 之ios 霍夫圓變換
目標(biāo)
在這個(gè)教程中你將學(xué)習(xí)如何:
- 使用OpenCV函數(shù) HoughCircles 在圖像中檢測圓.
原理
霍夫圓變換
霍夫圓變換的基本原理和上個(gè)教程中提到的霍夫線變換類似, 只是點(diǎn)對應(yīng)的二維極徑極角空間被三維的圓心點(diǎn)x, y還有半徑r空間取代.
-
對直線來說, 一條直線能由參數(shù)極徑極角 (r,θ)表示. 而對圓來說, 我們需要三個(gè)參數(shù)來表示一個(gè)圓, 如上文所說現(xiàn)在原圖像的邊緣圖像的任意點(diǎn)對應(yīng)的經(jīng)過這個(gè)點(diǎn)的所有可能圓是在三維空間有下面這三個(gè)參數(shù)來表示了，其對應(yīng)一條三維空間的曲線. 那么與二維的霍夫線變換同樣的道理, 對于多個(gè)邊緣點(diǎn)越多這些點(diǎn)對應(yīng)的三維空間曲線交于一點(diǎn)那么他們經(jīng)過的共同圓上的點(diǎn)就越多，類似的我們也就可以用同樣的閾值的方法來判斷一個(gè)圓是否被檢測到, 這就是標(biāo)準(zhǔn)霍夫圓變換的原理, 但也正是在三維空間的計(jì)算量大大增加的原因, 標(biāo)準(zhǔn)霍夫圓變化很難被應(yīng)用到實(shí)際中:
這里的(xcenter,ycenter)表示圓心的位置 (下圖中的綠點(diǎn)) 而r 表示半徑, 這樣我們就能唯一的定義一個(gè)圓了, 見下圖:
- 出于上面提到的對運(yùn)算效率的考慮, OpenCV實(shí)現(xiàn)的是一個(gè)比標(biāo)準(zhǔn)霍夫圓變換更為靈活的檢測方法: 霍夫梯度法, 也叫2-1霍夫變換(21HT), 它的原理依據(jù)是圓心一定是在圓上的每個(gè)點(diǎn)的模向量上, 這些圓上點(diǎn)模向量的交點(diǎn)就是圓心, 霍夫梯度法的第一步就是找到這些圓心, 這樣三維的累加平面就又轉(zhuǎn)化為二維累加平面. 第二部根據(jù)所有候選中心的邊緣非0像素對其的支持程度來確定半徑. 21HT方法最早在Illingworth的論文The Adaptive Hough Transform中提出并詳細(xì)描述, 也可參照Yuen在1990年發(fā)表的A Comparative Study of Hough Transform Methods for Circle Finding, Bradski的《學(xué)習(xí)OpenCV》一書則對OpenCV中具體對算法的具體實(shí)現(xiàn)有詳細(xì)描述并討論了霍夫梯度法的局限性.
例程
這個(gè)例程是用來干嘛的?
- 加載一幅圖像并對其模糊化以降噪
- 對模糊化后的圖像執(zhí)行霍夫圓變換 .
- 在窗體中顯示檢測到的圓.
#ifdef __cplusplus
#import <opencv2/opencv.hpp>
#import <opencv2/imgcodecs/ios.h>
#import <opencv2/imgproc.hpp>
#import <opencv2/highgui.hpp>
#import <opencv2/core/operations.hpp>
#import <opencv2/core/core_c.h>
using namespace cv;
using namespace std;
#endif
#import "HoughCirclesViewController.h"
@interface HoughCirclesViewController ()
@end
@implementation HoughCirclesViewController
const int cannyThresholdInitialValue = 100;
const int accumulatorThresholdInitialValue = 50;
const int maxAccumulatorThreshold = 200;
const int maxCannyThreshold = 255;
int cannyThreshold = cannyThresholdInitialValue;
int accumulatorThreshold = accumulatorThresholdInitialValue;
Mat src, src_gray;
- (void)viewDidLoad {
[super viewDidLoad];
UIImage * src1Image = [UIImage imageNamed:@"stuff.jpg"];
src = [self cvMatFromUIImage:src1Image];
UIImageView *imageView;
imageView = [self createImageViewInRect:CGRectMake(0, 100, 150, 150)];
[self.view addSubview:imageView];
imageView.image = [self UIImageFromCVMat:src];
cvtColor( src, src_gray, COLOR_BGR2GRAY );
imageView = [self createImageViewInRect:CGRectMake(0, 250, 150, 150)];
[self.view addSubview:imageView];
imageView.image = [self UIImageFromCVMat:src_gray];
GaussianBlur( src_gray, src_gray, cv::Size(9, 9), 2, 2 );
imageView = [self createImageViewInRect:CGRectMake(0, 400, 150, 150)];
[self.view addSubview:imageView];
imageView.image = [self UIImageFromCVMat:src_gray];
[self createSliderFrame:CGRectMake(150, 100, 100, 50) maxValue:maxCannyThreshold minValue:0 block:^(float value) {
cannyThreshold = std::max(cannyThreshold, 1);
[self HoughDetection];
}];
[self createSliderFrame:CGRectMake(150, 150, 100, 50) maxValue:maxAccumulatorThreshold minValue:0 block:^(float value) {
accumulatorThreshold = std::max(accumulatorThreshold, 1);
[self HoughDetection];
}];
[self HoughDetection];
}
-(void)HoughDetection
{
std::vector<Vec3f> circles;
// runs the actual detection
HoughCircles( src_gray, circles, HOUGH_GRADIENT, 1, src_gray.rows/8, cannyThreshold, accumulatorThreshold, 0, 0 );
// clone the colour, input image for displaying purposes
Mat display = src.clone();
for( size_t i = 0; i < circles.size(); i++ )
{
cv::Point center(cvRound(circles[i][0]), cvRound(circles[i][1]));
int radius = cvRound(circles[i][2]);
// circle center
circle( display, center, 3, Scalar(0,255,0), -1, 8, 0 );
// circle outline
circle( display, center, radius, Scalar(0,0,255), 3, 8, 0 );
}
UIImageView *imageView;
imageView = [self createImageViewInRect:CGRectMake(150, 250, 150, 150)];
[self.view addSubview:imageView];
imageView.image = [self UIImageFromCVMat:display];
}
#pragma mark - private
//brg
- (cv::Mat)cvMatFromUIImage:(UIImage *)image
{
CGColorSpaceRef colorSpace =CGColorSpaceCreateDeviceRGB();
CGFloat cols = image.size.width;
CGFloat rows = image.size.height;
Mat cvMat(rows, cols, CV_8UC4); // 8 bits per component, 4 channels (color channels + alpha)
CGContextRef contextRef = CGBitmapContextCreate(cvMat.data, // Pointer to data
cols, // Width of bitmap
rows, // Height of bitmap
8, // Bits per component
cvMat.step[0], // Bytes per row
colorSpace, // Colorspace
kCGImageAlphaNoneSkipLast |
kCGBitmapByteOrderDefault); // Bitmap info flags
CGContextDrawImage(contextRef, CGRectMake(0, 0, cols, rows), image.CGImage);
CGContextRelease(contextRef);
Mat dst;
Mat src;
cvtColor(cvMat, dst, COLOR_RGBA2BGRA);
cvtColor(dst, src, COLOR_BGRA2BGR);
return src;
}
-(UIImage *)UIImageFromCVMat:(cv::Mat)cvMat
{
// mat 是brg 而 rgb
Mat src;
NSData *data=nil;
CGBitmapInfo info =kCGImageAlphaNone|kCGBitmapByteOrderDefault;
CGColorSpaceRef colorSpace;
if (cvMat.depth()!=CV_8U) {
Mat result;
cvMat.convertTo(result, CV_8U,255.0);
cvMat = result;
}
if (cvMat.elemSize() == 1) {
colorSpace = CGColorSpaceCreateDeviceGray();
data= [NSData dataWithBytes:cvMat.data length:cvMat.elemSize()*cvMat.total()];
} else if(cvMat.elemSize() == 3){
cvtColor(cvMat, src, COLOR_BGR2RGB);
data= [NSData dataWithBytes:src.data length:src.elemSize()*src.total()];
colorSpace = CGColorSpaceCreateDeviceRGB();
}else if(cvMat.elemSize() == 4){
colorSpace = CGColorSpaceCreateDeviceRGB();
cvtColor(cvMat, src, COLOR_BGRA2RGBA);
data= [NSData dataWithBytes:src.data length:src.elemSize()*src.total()];
info =kCGImageAlphaNoneSkipLast | kCGBitmapByteOrderDefault;
}else{
NSLog(@"[error:] 錯(cuò)誤的顏色通道");
return nil;
}
CGDataProviderRef provider = CGDataProviderCreateWithCFData((__bridge CFDataRef)data);
// Creating CGImage from cv::Mat
CGImageRef imageRef = CGImageCreate(cvMat.cols, //width
cvMat.rows, //height
8, //bits per component
8 * cvMat.elemSize(), //bits per pixel
cvMat.step[0], //bytesPerRow
colorSpace, //colorspace
kCGImageAlphaNone|kCGBitmapByteOrderDefault,// bitmap info
provider, //CGDataProviderRef
NULL, //decode
false, //should interpolate
kCGRenderingIntentAbsoluteColorimetric //intent
);
// Getting UIImage from CGImage
UIImage *finalImage = [UIImage imageWithCGImage:imageRef];
CGImageRelease(imageRef);
CGDataProviderRelease(provider);
CGColorSpaceRelease(colorSpace);
return finalImage;
}
@end
解釋
- 執(zhí)行霍夫圓變換:
-(void)HoughDetection
{
std::vector<Vec3f> circles;
// runs the actual detection
HoughCircles( src_gray, circles, HOUGH_GRADIENT, 1, src_gray.rows/8, cannyThreshold, accumulatorThreshold, 0, 0 );
// clone the colour, input image for displaying purposes
Mat display = src.clone();
for( size_t i = 0; i < circles.size(); i++ )
{
cv::Point center(cvRound(circles[i][0]), cvRound(circles[i][1]));
int radius = cvRound(circles[i][2]);
// circle center
circle( display, center, 3, Scalar(0,255,0), -1, 8, 0 );
// circle outline
circle( display, center, radius, Scalar(0,0,255), 3, 8, 0 );
}
UIImageView *imageView;
imageView = [self createImageViewInRect:CGRectMake(150, 250, 150, 150)];
[self.view addSubview:imageView];
imageView.image = [self UIImageFromCVMat:display];
}
函數(shù)帶有以下自變量:
src_gray: 輸入圖像 (灰度圖)
-
circles: 存儲(chǔ)下面三個(gè)參數(shù):xc,:yc,r
集合的容器來表示每個(gè)檢測到的圓.
CV_HOUGH_GRADIENT: 指定檢測方法. 現(xiàn)在OpenCV中只有霍夫梯度法
dp = 1: 累加器圖像的反比分辨率
min_dist = src_gray.rows/8: 檢測到圓心之間的最小距離
param_1 = 200: Canny邊緣函數(shù)的高閾值
param_2 = 100: 圓心檢測閾值.
min_radius = 0: 能檢測到的最小圓半徑, 默認(rèn)為0.
max_radius = 0: 能檢測到的最大圓半徑, 默認(rèn)為0
結(jié)果
使用圖片
smarties.png
