边缘检测类(包括Roberts, Sobel, Prewitt, Kirsch等算子的边缘检测算法) - wtuyzh的专栏 - CSDNBlog

        边缘检测类(包括Roberts, Sobel, Prewitt, Kirsch等算子的边缘检测算法)
public class EdgeDetect : ImageInfo
{
/************************************************************
*
* Roberts, Sobel, Prewitt, Kirsch, GaussLaplacian
* 水平检测、垂直检测、边缘增强、边缘均衡化
*
************************************************************/


///
/// 对两幅图像进行梯度运算
///
/// 位图 1
/// 位图 2
///
private Bitmap Gradient(Bitmap b1, Bitmap b2)
{

int width = b1.Width;
int height = b1.Height;
BitmapData data1 = b1.LockBits(new Rectangle(0, 0, width, height),
ImageLockMode.ReadWrite, PixelFormat.Format32bppArgb);
BitmapData data2 = b2.LockBits(new Rectangle(0, 0, width, height),
ImageLockMode.ReadOnly, PixelFormat.Format32bppArgb);
unsafe
{
byte* p1 = (byte*)data1.Scan0;
byte* p2 = (byte*)data2.Scan0;
int offset = data1.Stride - width * BPP;
for (int y = 0; y < height; y++)
{
for (int x = 0; x < width; x++)
{
for (int i = 0; i < 3; i++)
{
int power = (int)Math.Sqrt((p1[i] * p1[i] + p2[i] * p2[i]));
p1[i] = (byte)(power > 255 ? 255 : power);
} // i
p1 += BPP;
p2 += BPP;
} // x
p1 += offset;
p2 += offset;
} // y
}
b1.UnlockBits(data1);
b2.UnlockBits(data2);
Bitmap dstImage = (Bitmap)b1.Clone();
b1.Dispose();
b2.Dispose();
return dstImage;
} // end of Gradient

///
/// 按 Roberts 算子进行边缘检测
///
/// 位图流
///
public Bitmap Roberts(Bitmap b)
{
int width = b.Width;
int height = b.Height;
Bitmap dstImage = new Bitmap(width, height);
BitmapData srcData = b.LockBits(new Rectangle(0, 0, width, height),
ImageLockMode.ReadOnly, PixelFormat.Format32bppArgb);
BitmapData dstData = dstImage.LockBits(new Rectangle(0, 0, width, height),
ImageLockMode.WriteOnly, PixelFormat.Format32bppArgb);
int stride = srcData.Stride;
int offset = stride - width * BPP;
unsafe
{
byte* src = (byte*)srcData.Scan0;
byte* dst = (byte*)dstData.Scan0;
int A, B; // A(x-1, y-1) B(x, y-1)
int C, D; // C(x-1, y) D(x, y)
// 指向第一行
src += stride;
dst += stride;
// 不处理最上边和最左边
for (int y = 1; y < height; y++)
{
// 指向每行第一列
src += BPP;
dst += BPP;
for (int x = 1; x < width; x++)
{
for (int i = 0; i < 3; i++)
{
A = src[i - stride - BPP];
B = src[i - stride];
C = src[i - BPP];
D = src[i];
dst[i] = (byte)(Math.Sqrt((A - D) * (A - D) + (B - C) * (B - C)));
} // i
dst[3] = src[3];
src += BPP;
dst += BPP;
} // x
src += offset;
dst += offset;
} // y
}
b.UnlockBits(srcData);
dstImage.UnlockBits(dstData);
b.Dispose();
return dstImage;
} // end of Roberts

/// /// 按 Sobel 算子进行边缘检测
///
/// 位图流
///
public Bitmap Sobel(Bitmap b)
{
Matrix3x3 m = new Matrix3x3();
// -1 -2 -1
// 0 0 0
// 1 2 1
m.Init(0);
m.TopLeft = m.TopRight = -1;
m.BottomLeft = m.BottomRight = 1;
m.TopMid = -2;
m.BottomMid = 2;
Bitmap b1 = m.Convolute((Bitmap)b.Clone());
// -1 0 1
// -2 0 2
// -1 0 1
m.Init(0);
m.TopLeft = m.BottomLeft = -1;
m.TopRight = m.BottomRight = 1;
m.MidLeft = -2;
m.MidRight = 2;
Bitmap b2 = m.Convolute((Bitmap)b.Clone());
// 0 1 2
// -1 0 1
// -2 -1 0
m.Init(0);
m.TopMid = m.MidRight = 1;
m.MidLeft = m.BottomMid = -1;
m.TopRight = 2;
m.BottomLeft = -2;
Bitmap b3 = m.Convolute((Bitmap)b.Clone());
// -2 -1 0
// -1 0 1
// 0 1 2
m.Init(0);
m.TopMid = m.MidLeft = -1;
m.MidRight = m.BottomMid = 1;
m.TopLeft = -2;
m.BottomRight = 2;
Bitmap b4 = m.Convolute((Bitmap)b.Clone());
// 梯度运算
b = Gradient(Gradient(b1, b2), Gradient(b3, b4));
b1.Dispose(); b2.Dispose(); b3.Dispose(); b4.Dispose();
return b;
} // end of Sobel

///
/// 按 Prewitt 算子进行边缘检测
///
/// 位图流
///
public Bitmap Prewitt(Bitmap b)
{
Matrix3x3 m = new Matrix3x3();
// -1 -1 -1
// 0 0 0
// 1 1 1
m.Init(0);
m.TopLeft = m.TopMid = m.TopRight = -1;
m.BottomLeft = m.BottomMid = m.BottomRight = 1;
Bitmap b1 = m.Convolute((Bitmap)b.Clone());
// -1 0 1
// -1 0 1
// -1 0 1
m.Init(0);
m.TopLeft = m.MidLeft = m.BottomLeft = -1;
m.TopRight = m.MidRight = m.BottomRight = 1;
Bitmap b2 = m.Convolute((Bitmap)b.Clone());
// -1 -1 0
// -1 0 1
// 0 1 1
m.Init(0);
m.TopLeft = m.MidLeft = m.TopMid = -1;
m.BottomMid = m.BottomRight = m.MidRight = 1;
Bitmap b3 = m.Convolute((Bitmap)b.Clone());
// 0 1 1
// -1 0 1
// -1 -1 0
m.Init(0);
m.TopMid = m.TopRight = m.MidRight = 1;
m.MidLeft = m.BottomLeft = m.BottomMid = -1;
Bitmap b4 = m.Convolute((Bitmap)b.Clone());
// 梯度运算
b = Gradient(Gradient(b1, b2), Gradient(b3, b4));
b1.Dispose(); b2.Dispose(); b3.Dispose(); b4.Dispose();
return b;
} // end of Prewitt

///
/// 按 Kirsch 算子进行边缘检测
///
/// 位图流
///
public Bitmap Kirsch(Bitmap b)
{
Matrix3x3 m = new Matrix3x3();
// 5 5 5
// -3 0 -3
// -3 -3 -3
m.Init(-3);
m.Center = 0;
m.TopLeft = m.TopMid = m.TopRight = 5;
Bitmap b1 = m.Convolute((Bitmap)b.Clone());
// -3 5 5
// -3 0 5
// -3 -3 -3
m.Init(-3);
m.Center = 0;
m.TopMid = m.TopRight = m.MidRight = 5;
Bitmap b2 = m.Convolute((Bitmap)b.Clone());
// -3 -3 5
// -3 0 5
// -3 -3 5
m.Init(-3);
m.Center = 0;
m.TopRight = m.MidRight = m.BottomRight = 5;
Bitmap b3 = m.Convolute((Bitmap)b.Clone());
// -3 -3 -3
// -3 0 5
// -3 5 5
m.Init(-3);
m.Center = 0;
m.MidRight = m.BottomRight = m.BottomMid = 5;
Bitmap b4 = m.Convolute((Bitmap)b.Clone());
// -3 -3 -3
// -3 0 -3
// 5 5 5
m.Init(-3);
m.Center = 0;
m.BottomLeft = m.BottomMid = m.BottomRight = 5;
Bitmap b5 = m.Convolute((Bitmap)b.Clone());
// -3 -3 -3
// 5 0 -3
// 5 5 -3
m.Init(-3);
m.Center = 0;
m.MidLeft = m.BottomLeft = m.BottomMid = 5;
Bitmap b6 = m.Convolute((Bitmap)b.Clone());
// 5 -3 -3
// 5 0 -3
// 5 -3 -3
m.Init(-3);
m.Center = 0;
m.TopLeft = m.MidLeft = m.BottomLeft = 5;
Bitmap b7 = m.Convolute((Bitmap)b.Clone());
// 5 5 -3
// 5 0 -3
// -3 -3 -3
m.Init(-3);
m.Center = 0;
m.TopLeft = m.MidLeft = m.TopMid = 5;
Bitmap b8 = m.Convolute((Bitmap)b.Clone());
// 梯度运算
Bitmap b9 = Gradient(Gradient(b1, b2), Gradient(b3, b4));
Bitmap b10 = Gradient(Gradient(b5, b6), Gradient(b7, b8));
b = Gradient(b9, b10);
b1.Dispose(); b2.Dispose(); b3.Dispose(); b4.Dispose();
b5.Dispose(); b6.Dispose(); b7.Dispose(); b8.Dispose();
b9.Dispose(); b10.Dispose();
return b;
} // end of Kirsch

///
/// 按 GaussLaplacian 算子进行边缘检测
///
///
///
public Bitmap GaussLaplacian(Bitmap b)
{
int[,] kernel = {
{-2, -4, -4, -4, -2},
{-4, 0, 8, 0, -4},
{-4, 8, 24, 8, -4},
{-4, 0, 8, 0, -4},
{-2, -4, -4, -4, -2}
};
MatrixNxN m = new MatrixNxN();
m.Kernel = kernel;
return m.Convolute(b);
} // end of GaussLaplacian

///
/// 按水平边缘检测算子进行边缘检测
///
///
///
public Bitmap EdgeDetectHorizontal(Bitmap b)
{
int[,] kernel = {
{ 0, 0, 0, 0, 0, 0, 0},
{ 0, 0, 0, 0, 0, 0, 0},
{-1, -1, -1, -1, -1, -1, -1},
{ 0, 0, 0, 0, 0, 0, 0},
{ 1, 1, 1, 1, 1, 1, 1},
{ 0, 0, 0, 0, 0, 0, 0},
{ 0, 0, 0, 0, 0, 0, 0},
};
MatrixNxN m = new MatrixNxN();
m.Kernel = kernel;
return m.Convolute(b);
} // end of EdgeDetectHorizontal

///
/// 按垂直边缘检测算子进行边缘检测
///
///
///
public Bitmap EdgeDetectVertical(Bitmap b)
{
int[,] kernel = {
{ 0, 0, -1, 0, 1, 0, 0},
{ 0, 0, -1, 0, 1, 0, 0},
{ 0, 0, -1, 0, 1, 0, 0},
{ 0, 0, -1, 0, 1, 0, 0},
{ 0, 0, -1, 0, 1, 0, 0},
{ 0, 0, -1, 0, 1, 0, 0},
{ 0, 0, -1, 0, 1, 0, 0},
};
MatrixNxN m = new MatrixNxN();
m.Kernel = kernel;
return m.Convolute(b);
} // end of EdgeDetectVertical

///
/// 边缘增强
///
/// 位图流
/// 阈值
///
public Bitmap EdgeEnhance(Bitmap b, int threshold)
{
int width = b.Width;
int height = b.Height;
Bitmap dstImage = (Bitmap)b.Clone();
BitmapData srcData = b.LockBits(new Rectangle(0, 0, width, height),
ImageLockMode.ReadOnly, PixelFormat.Format32bppArgb);
BitmapData dstData = dstImage.LockBits(new Rectangle(0, 0, width, height),
ImageLockMode.WriteOnly, PixelFormat.Format32bppArgb);
// 图像实际处理区域
// 不考虑最左 1 列和最右 1 列
// 不考虑最上 1 行和最下 1 行
int rectTop = 1;
int rectBottom = height - 1;
int rectLeft = 1;
int rectRight = width - 1;
unsafe
{
byte* src = (byte*)srcData.Scan0;
byte* dst = (byte*)dstData.Scan0;
int stride = srcData.Stride;
int offset = stride - width * BPP;
int pixel = 0;
int maxPixel = 0;

// 指向第 1 行
src += stride;
dst += stride;
for (int y = rectTop; y < rectBottom; y++)
{
// 指向每行第 1 列像素
src += BPP;
dst += BPP;
for (int x = rectLeft; x < rectRight; x++)
{
// Alpha
dst[3] = src[3];
// 处理 B, G, R 三分量
for (int i = 0; i < 3; i++)
{
// 右上-左下
maxPixel = src[i - stride + BPP] - src[i + stride - BPP];
if (maxPixel < 0) maxPixel = -maxPixel;
// 左上-右下
pixel = src[i - stride - BPP] - src[i + stride + BPP];
if (pixel < 0) pixel = -pixel;
if (pixel > maxPixel) maxPixel = pixel;
// 上-下
pixel = src[i - stride] - src[i + stride];
if (pixel < 0) pixel = -pixel;
if (pixel > maxPixel) maxPixel = pixel;
// 左-右
pixel = src[i - BPP] - src[i + BPP];
if (pixel < 0) pixel = -pixel;
if (pixel > maxPixel) maxPixel = pixel;
// 进行阈值判断
if (maxPixel < threshold) maxPixel = 0;

dst[i] = (byte)maxPixel;
}
// 向后移一像素
src += BPP;
dst += BPP;
} // x
// 移向下一行
// 这里得注意要多移 1 列，因最右边还有 1 列不必处理
src += offset + BPP;
dst += offset + BPP;
} // y
}
// b.UnlockBits(srcData);
dstImage.UnlockBits(dstData);
b.Dispose();
return dstImage;
} // end of EdgeEnhance



}


本文转自
http://blog.csdn.net/wtuyzh/archive/2006/07/05/881518.aspx