15,917,859 members
Articles / Programming Languages / C++
Tip/Trick

# Detect Amount of Arc

Rate me:
11 Dec 2013CPOL1 min read 14.2K   396   4   3
With this algorithm, we can detect amount and knoll of an arc

## Introduction

Today, Image processing does a lot of work and can be applied in every subject such as medic, tumor detection, driving, speed getting, traffic analyse, edge detection, background extraction, text extraction, ...

With this algorithm, we can find knoll and get amount of arc of a line in every picture easily.

## Using the Code And Algorithm

First of all, we must load the image from hard to program. We do it with `CImage `Object.

C++
```CImage i;

After that, we should grey the image for more easy processing. Working with two colors, black and white. We put black color where RGB is less than 80 and put white where RGB is more than 80 value. To convert the image to gray-mode:

C++
```CDC cdc;//declaration an object from Class Device Content
cdc.Attach(i.GetDC());//get Device Content from CImage (i)
COLORREF c,c1;//define two color
long r,g,b;//define red ,green, blue values
int avg;
for(int y=0;y<i.GetHeight();y++)
for(int x=0;x<i.GetWidth();x++){
c=cdc.GetPixel(x,y);//get color from image
//get RGB
r=GetRValue(c); g=GetGValue(c);  b=GetBValue(c); avg=(r+g+b)/3;
if(avg>80)r=255;
else r=0;
cdc.SetPixel(x,y,RGB(r,r,r));//put color to image
}
cdc.Detach();
i.ReleaseDC();```

The next step is arc detection. The algorithm is:

First we look for first pixel that is not the back color (white) and call `Point 1 (x1,y1)`. Then continue. Next pixel that has this trait, if it’s distance from pre point, it belongs to arc, else not. Keep on until we won't find the pixel that belongs to arc then last point we call `Point 2(x2,y2)`. `Point 3(x3,y3)` is knoll of arc. Put `(x3,y3)` to pixel that has higher height.

C++
```CDC cdc;
cdc.Attach(i.GetDC());
CPen p(PS_SOLID,1,RGB(255,255,0)),p1(PS_SOLID,1,RGB(0,0,255)),p2(PS_DASH,1,RGB(255,0,0));
cdc.SelectObject(&p);
register int x1=-1,x2,x3,y1,y2,y3,d=10;x1=x2=x3=y1=y2=y3=-1;
register COLORREF c;
for(register int x=0;x<i.GetWidth();x++)

for(register int y=0;y<i.GetHeight();y++){

c=i.GetPixel(x,y);
if(c!=RGB(255,255,255)){//line detect
if(x1==-1){//first point of line
x1=x2=x3=x;y1=y2=y3=y;
}else if(sqrt(pow(x-x2,2.0)+pow(y-y2,2.0))<=d){//next point
x2=x;y2=y;
if(abs(y2-y1)>=abs(y3-y1)){
x3=x2;y3=y2;
}
cdc.SetPixel(x,y,RGB(255,255,0));
}
}
}
register float length=sqrt(pow(x1-x2,2.0)+pow(y1-y2,2.0));
cdc.SelectObject(&p1);
cdc.MoveTo(x3,y1);cdc.LineTo(x3,y3);
cdc.SelectObject(&p2);cdc.MoveTo(x1,y1);cdc.LineTo(x2,y2);
cdc.Detach();i.ReleaseDC();OnPaint();
float arc=4000*abs(y3-y2)/length;
CString s;char a[20]="";itoa(arc,a,10);s=a;s+=_T(" mM");	```

1. Original Image

2. Grey image

3. Processed image

## The Solution And Points

For detecting the knoll of an arc without its function is difficult and we must use digital analyse to do this. But with image processing and this algorithm, we can do it easily and fast. It's useful.

Written By
Software Developer
Iran (Islamic Republic of)
This member has not yet provided a Biography. Assume it's interesting and varied, and probably something to do with programming.

 C++ `if(abs(y2-y1)>=abs(y3-y1))` This is a strange way to find a maximum (assuming the "knoll" is the maximum of the curve). The condition can return true for a maximum, but also a minimum or an inflection point. And if there are two maxima or more, it is unsure which one will be reported.