OpenCV指针仪表盘参数读取（四）指针定位

这一节是此方案设计的最核心部分&＃xff0c;指针能否准确的定位出来是最关键的问题&＃xff0c;在学习过程中&＃xff0c;初步采用几种方案来尽可能的提高指针定位的精度&＃xff1a;

1.指针定位的方案&＃xff1a; Hough直线检测 环向模板匹配法 径向灰度求和法

2.指针细化的方法&＃xff08;算法将放在文章最后提供测试与使用&＃xff09;

对于指针定位&＃xff0c;初步的设计思路是采用Hough直线检测的方法&＃xff0c;而后通过直线上的点与圆的关系来将错误检测的直线进行过滤&＃xff0c;此方案的难点在于Hough直线检测的方法经常与Canny边缘检测算子进行联合使用&＃xff0c;而Canny算子的阈值选择以及Hough直线的长度选择都是比较动态的过程&＃xff0c;在调节过程非常困难&＃xff0c;本文采用可以连续调节Canny算子和Hough检测阈值的方法&＃xff0c;来选择适合于方案的值&＃xff0c;初步测试有较好的效果。

函数使用如下&＃xff1a;

// HoughLine
midd_line_img &＃61; HoughLine(midd_img, midd_line_img);

OpenCV中的函数封装为:

// CannyDetect
Canny(gray_img, midd_line_img, 23, 55, 3);
// HoughLine
HoughLinesP(midd_line_img, mylines, 1, CV_PI / 180, g_nthreshold &＃43; 1, 20, 5);

采用的可以调节的Canny算子的程序及效果图如下&＃xff1a;

#include
#include
#include
using namespace std;
using namespace cv;
// Define the Mat parameters
Mat g_srcImage, g_dstImage, g_midImage;
// Define a vector to collect all detected lines
vector g_lines;
int g_nthreshold &＃61; 100;
static void showImage1(int, void*);
static void showImage2(int, void*);
int thred1 &＃61; 23;
int thred2 &＃61; 55;
// 23 55
int main()
{
// Load origin image
g_srcImage &＃61; imread("12.jpg", 0);
// Show origin image
namedWindow("Dst Image", 1);
imshow("Origin Image", g_srcImage);
// Create a Tracebar
createTrackbar("threshold1", "Dst Image", &thred1, 200, showImage1);
createTrackbar("threshold2", "Dst Image", &thred2, 200, showImage2);
// Canny Detect and Gray am image
Canny(g_srcImage, g_midImage, thred1, thred2, 3);
cvtColor(g_midImage, g_dstImage, CV_GRAY2BGR);
showImage1(thred1, 0);
showImage2(thred2, 0);
// Show the dst image
imshow("Dst Image", g_dstImage);
waitKey(0);
return 0;
}
static void showImage1(int thred1, void*) {
// Canny Detection
Canny(g_srcImage, g_midImage, thred1, thred2, 3);
imshow("Dst Image", g_midImage);
}
static void showImage2(int thred2, void*) {
// Canny Detection
Canny(g_srcImage, g_midImage, thred1, thred2, 3);
imshow("Dst Image", g_midImage);
}

效果图&＃xff1a;

原图&＃xff1a;

适宜的阈值选择&＃xff1a;

不适宜的阈值选择&＃xff1a;已经将指针直线湮没

同理&＃xff0c;HoughLine方法的总体方案如下&＃xff1a;

Mat HoughLine(Mat midd_img, Mat midd_line_img) {

Mat dst_img, gray_img;
cvtColor(midd_img, gray_img, COLOR_BGR2GRAY);
// CannyDetect
Canny(gray_img, midd_line_img, 23, 55, 3);
// convert cannied image to a gray one
cvtColor(midd_line_img, dst_img, CV_GRAY2BGR);
// define a vector to collect all possible lines
vector mylines;
int g_nthreshold &＃61; 39;
HoughLinesP(midd_line_img, mylines, 1, CV_PI / 180, g_nthreshold &＃43; 1, 20, 5);
// draw every line by using for
for (size_t i &＃61; 0; i {
l &＃61; mylines[i];
if (((circle_center.x - 10) <&＃61; l[0]) && (l[0] <&＃61; (circle_center.x &＃43; 10)))
if (((circle_center.y - 10) <&＃61; l[1]) && (l[1] <&＃61; (circle_center.y &＃43; 10)))
if (((circle_center.x - circle_radius) <&＃61; l[2]) && (l[2] <&＃61; (circle_center.x &＃43; circle_radius)))
if (((circle_center.y - circle_radius) <&＃61; l[3]) && (l[3] <&＃61; (circle_center.y &＃43; circle_radius)))
{
//cout < // line(midd_img, Point(l[0], l[1]), Point(l[2], l[3]), Scalar(23, 180, 55), 2, CV_AA);
cho_l &＃61; l;
line(midd_img, Point(circle_center.x, circle_center.y), Point(cho_l[2], cho_l[3]), Scalar(23, 180, 55), 2, CV_AA);
}
}
return midd_img;
}


效果图如下&＃xff1a;

欢迎评论文章&＃xff0c;给予宝贵建议&＃xff0c;工程提供在github下载学习&＃xff0c;请关注作者&＃xff0c;共同学习交流。