Image Segmentation--global threshold processing using iterative algorithm

Basic Global Threshold method

When the image is composed of a dark background and a lighter object, the method of extracting the object from the background is to select a threshold T that separates the two gray values, that is, any point of f (x, Y) > T (x, y) is called

As an object point, otherwise the point is called a background point. Then, the segmented image g (x, Y) is given by the following formula:

Over here:

The threshold t can be a constant, then the global threshold processing is used

The threshold T can be changed on a graph, and a variable threshold is used to process

Threshold T can be multiple, use double (multi) threshold processing

For example, the following grayscale histogram, the left can use a single threshold, the right side requires a double threshold to split.

global threshold processing with an iterative approach

In real-world processing, there is usually a big change between images, even if it is possible to use global thresholds for each picture, but we also need to be able to self-

The threshold value of the motion and estimation. Therefore, the iterative approach is used for this purpose:

1. Select an initial value for the global threshold T (average grayscale selection)

2. Split the image with T, which produces two sets of pixels: G1 consists of all pixels with a grayscale value greater than T, G2 by a grayscale value less than t

Make up all the pixels

3. Calculate average gray values M1 and M2 for G1 and G2 pixels respectively

4. Calculate a new threshold value:

t= (M1+M2)/2

5. Repeat steps 2 through 4 until the difference between T values in the continuous iteration is less than a predefined parameter T.

Nmaxiter: Maximum number of iterations; Ndiffrec: Average gray difference between bright and dark regions using given threshold value  
 int detectthreshold (iplimage*img, int nmaxiter, int& IDIFFREC)//Threshold Segmentation: Iterative method  
 {  
 		 //image information  
		 int height = img->height;  
		 int width = img->width;  
		 int step = img->widthstep/sizeof (Uchar);  
		 Uchar *data = (uchar*) img->imagedata;  
		   
		 Idiffrec =0;  
		 int f[256]={0}; The histogram array  
		 int itotalgray=0;//The grayscale value and  
		 int itotalpixel =0;//  
		 pixels and the pixel value of a byte bt;//point  
		   
		 uchar Ithrehold, inewthrehold;//threshold, New threshold  
		 Uchar imaxgrayvalue=0,imingrayvalue=255;//maximum grayscale value and minimum grayscale value in the original image  
		 Uchar iMeanGrayValue1 , iMeanGrayValue2;  
		   
		 Gets the value of (i,j), stored in the histogram array F for  
		 (int i=0;iimaxgrayvalue)  
				 		imaxgrayvalue = BT;  
				 f[bt]++;  
			 }  
		 }  
		   
		 Ithrehold =0;//  
		 inewthrehold = (imingrayvalue+imaxgrayvalue)/2;//initial threshold  
		 Idiffrec = imaxgrayvalue- Imingrayvalue;  
		   
		 for (int a=0; (ABS (Ithrehold-inewthrehold) >0.5) &&a