使用C與C++/CLI實現Sobel Edge Detector。

http://www.cnblogs.com/oomusou/archive/2008/07/23/sobel_edge_detector.html

 /*
(C) OOMusou 2007 http://oomusou.cnblogs.com Filename : sobel_edge.c
Compiler : Visual C++ 8.0
Description : Demo the how to use sobel detector on gray level image
Release : 07/23/2008 1.0
*/
#include <stdio.h>
#include <stdlib.h>
#include <math.h> #define MASK_N 2
#define MASK_X 3
#define MASK_Y 3
#define WHITE 255
#define BLACK 0 unsigned char *image_s = NULL; // source image array
unsigned char *image_t = NULL; // target image array
FILE *fp_s = NULL; // source file handler
FILE *fp_t = NULL; // target file handler unsigned int width, height; // image width, image height
unsigned int rgb_raw_data_offset;// RGB raw data offset
unsigned char bit_per_pixel; // bit per pixel
unsigned short byte_per_pixel; // byte per pixel // bitmap header
unsigned char header[] = {
0x42, // identity : B
0x4d, // identity : M
, , , , // file size
, , // reserved1
, , // reserved2
, , , , // RGB data offset
, , , , // struct BITMAPINFOHEADER size
, , , , // bmp width
, , , , // bmp height
, , // planes
, , // bit per pixel
, , , , // compression
, , , , // data size
, , , , // h resolution
, , , , // v resolution
, , , , // used colors
, , , // important colors
}; // sobel mask
int mask[MASK_N][MASK_X][MASK_Y] = {
{{-,-,-},
{ , , },
{ , , }}, {{-, , },
{-, , },
{-, , }}
}; int read_bmp(const char *fname_s) {
fp_s = fopen(fname_s, "rb");
if (fp_s == NULL) {
printf("fopen fp_s error\n");
return -;
} // move offset to 10 to find rgb raw data offset
fseek(fp_s, , SEEK_SET);
fread(&rgb_raw_data_offset, sizeof(unsigned int), , fp_s); // move offset to 18 to get width & height;
fseek(fp_s, , SEEK_SET);
fread(&width, sizeof(unsigned int), , fp_s);
fread(&height, sizeof(unsigned int), , fp_s); // get bit per pixel
fseek(fp_s, , SEEK_SET);
fread(&bit_per_pixel, sizeof(unsigned short), , fp_s);
byte_per_pixel = bit_per_pixel / ; // move offset to rgb_raw_data_offset to get RGB raw data
fseek(fp_s, rgb_raw_data_offset, SEEK_SET); image_s = (unsigned char *)malloc((size_t)width * height * byte_per_pixel);
if (image_s == NULL) {
printf("malloc images_s error\n");
return -;
} image_t = (unsigned char *)malloc((size_t)width * height * byte_per_pixel);
if (image_t == NULL) {
printf("malloc image_t error\n");
return -;
} fread(image_s, sizeof(unsigned char), (size_t)(long)width * height * byte_per_pixel, fp_s); return ;
} // convert RGB to gray level int
int color_to_int(int r, int g, int b) {
return (r + g + b) / ;
} int sobel(double threshold) {
unsigned int x, y, i, v, u; // for loop counter
unsigned char R, G, B; // color of R, G, B
double val[MASK_N] = {0.0};
int adjustX, adjustY, xBound, yBound;
double total; for(y = ; y != height; ++y) {
for(x = ; x != width; ++x) {
for(i = ; i != MASK_N; ++i) {
adjustX = (MASK_X % ) ? : ;
adjustY = (MASK_Y % ) ? : ;
xBound = MASK_X / ;
yBound = MASK_Y / ; val[i] = 0.0;
for(v = -yBound; v != yBound + adjustY; ++v) {
for (u = -xBound; u != xBound + adjustX; ++u) {
if (x + u >= && x + u < width && y + v >= && y + v < height) {
R = *(image_s + byte_per_pixel * (width * (y+v) + (x+u)) + );
G = *(image_s + byte_per_pixel * (width * (y+v) + (x+u)) + );
B = *(image_s + byte_per_pixel * (width * (y+v) + (x+u)) + ); val[i] += color_to_int(R, G, B) * mask[i][u + xBound][v + yBound];
}
}
}
} total = 0.0;
for (i = ; i != MASK_N; ++i) {
total += val[i] * val[i];
} total = sqrt(total); if (total - threshold >= ) {
// black
*(image_t + byte_per_pixel * (width * y + x) + ) = BLACK;
*(image_t + byte_per_pixel * (width * y + x) + ) = BLACK;
*(image_t + byte_per_pixel * (width * y + x) + ) = BLACK;
}
else {
// white
*(image_t + byte_per_pixel * (width * y + x) + ) = WHITE;
*(image_t + byte_per_pixel * (width * y + x) + ) = WHITE;
*(image_t + byte_per_pixel * (width * y + x) + ) = WHITE;
}
}
} return ;
} int write_bmp(const char *fname_t) {
unsigned int file_size; // file size fp_t = fopen(fname_t, "wb");
if (fp_t == NULL) {
printf("fopen fname_t error\n");
return -;
} // file size
file_size = width * height * byte_per_pixel + rgb_raw_data_offset;
header[] = (unsigned char)(file_size & 0x000000ff);
header[] = (file_size >> ) & 0x000000ff;
header[] = (file_size >> ) & 0x000000ff;
header[] = (file_size >> ) & 0x000000ff; // width
header[] = width & 0x000000ff;
header[] = (width >> ) & 0x000000ff;
header[] = (width >> ) & 0x000000ff;
header[] = (width >> ) & 0x000000ff; // height
header[] = height &0x000000ff;
header[] = (height >> ) & 0x000000ff;
header[] = (height >> ) & 0x000000ff;
header[] = (height >> ) & 0x000000ff; // bit per pixel
header[] = bit_per_pixel; // write header
fwrite(header, sizeof(unsigned char), rgb_raw_data_offset, fp_t); // write image
fwrite(image_t, sizeof(unsigned char), (size_t)(long)width * height * byte_per_pixel, fp_t); fclose(fp_s);
fclose(fp_t); return ;
} int main() {
read_bmp("lena.bmp"); // 24 bit gray level image
sobel(90.0);
write_bmp("lena_sobel.bmp");
}
 /*
(C) OOMusou 2007 http://oomusou.cnblogs.com Filename : sobel_edge.cpp
Compiler : C++/CLI / Visual C++ 8.0
Description : Demo the how to use sobel detector on gray level image
Release : 07/23/2008 1.0
*/ #include "stdafx.h"
#include <cmath> using namespace System::Drawing;
using namespace System::Drawing::Imaging; const int MASK_N = ;
const int MASK_X = ;
const int MASK_Y = ; // Convert RGB to gray level int
int colorToInt(Color %color) {
return (color.R + color.G + color.B) / ;
} void edgeDetector(Bitmap^ oriImg, Bitmap^ resImg, const int mask[MASK_N][MASK_X][MASK_Y], const double &threshold) {
double val[MASK_N] = {0.0}; for(int y = ; y != oriImg->Height; ++y) {
for(int x = ; x != oriImg->Width; ++x) {
for(int i = ; i != MASK_N; ++i) {
int adjustX = (MASK_X % ) ? : ;
int adjustY = (MASK_Y % ) ? : ;
int xBound = MASK_X / ;
int yBound = MASK_Y / ; val[i] = 0.0;
for(int v = -yBound; v != yBound + adjustY; ++v) {
for (int u = -xBound; u != xBound + adjustX; ++u) {
if (x + u >= && x + u < oriImg->Width && y + v >= && y + v < oriImg->Height) {
val[i] += colorToInt(oriImg->GetPixel(x + u, y + v)) * mask[i][u + xBound][v + yBound];
}
}
}
} double total = 0.0;
for (int i = ; i != MASK_N; ++i) {
total += val[i] * val[i];
} total = sqrt(total); if (total - threshold >= )
resImg->SetPixel(x , y, Color::Black);
else
resImg->SetPixel(x, y, Color::White);
}
}
} int main() {
const int mask[MASK_N][MASK_X][MASK_Y] = {
{{-,-,-},
{ , , },
{ , , }}, {{-, , },
{-, , },
{-, , }}
}; Bitmap^ oriImg = gcnew Bitmap("lena.bmp");
Bitmap^ resImg = gcnew Bitmap(oriImg->Width, oriImg->Height); const double threshold = 90.0;
edgeDetector(oriImg, resImg, mask, threshold); resImg->Save("lena_sobel.bmp"); return ;
}

每天进步一点点------如何实现Sobel Edge Detector? (Image Processing) (C/C++)的更多相关文章

  1. Canny Edge Detector

    Canny边缘检测算法有自己的理论和经验性的推导, 没仔细看/没看明白. 它的步骤如下: 对原图的灰度图进行高斯滤波 求一阶导数, 得到每个像素点的梯度强度和方向. 非最大抑制. 对每个edge ca ...

  2. Understanding Convolution in Deep Learning

    Understanding Convolution in Deep Learning Convolution is probably the most important concept in dee ...

  3. 32个最热CPLD-FPGA论坛

    1. OPENCORES.ORG这里提供非常多,非常好的PLD了内核,8051内核就可以在里面找到.进入后,选择project或者由http//www.opencores.org/browse.cgi ...

  4. [转]FPGA网站推荐

    1. OPENCORES.ORG这里提供非常多,非常好的PLD了内核,8051内核就可以在里面找到.进入后,选择project或者由http//www.opencores.org/browse.cgi ...

  5. Computer Vision_33_SIFT:Remote Sensing Image Registration With Modified SIFT and Enhanced Feature Matching——2017

    此部分是计算机视觉部分,主要侧重在底层特征提取,视频分析,跟踪,目标检测和识别方面等方面.对于自己不太熟悉的领域比如摄像机标定和立体视觉,仅仅列出上google上引用次数比较多的文献.有一些刚刚出版的 ...

  6. OpenCV2马拉松第14圈——边缘检測(Sobel,prewitt,roberts)

    收入囊中 差分在边缘检測的角色 Sobel算子 OpenCV sobel函数 OpenCV Scharr函数 prewitt算子 Roberts算子 葵花宝典 差分在边缘检測究竟有什么用呢?先看以下的 ...

  7. Sobel Derivatives

    https://docs.opencv.org/2.4/doc/tutorials/imgproc/imgtrans/sobel_derivatives/sobel_derivatives.html ...

  8. 线特征---Edge Drawing(七)

    http://ceng.anadolu.edu.tr/cv/edgedrawing/ References C. Topal, C. Akinlar, Edge Drawing: A Combined ...

  9. 2.7 Sobel导数

    OpenCV函数 Sobel(src_gray,grad_x/grad_y,ddepth,x_order,y_order,scale,delta,BORDER_DEFAULT ) Scharr( ) ...

随机推荐

  1. How to write a paper in a weekend - by Prof. Pete Carr

    Key points: don't procrastinate; review the notes and renew the literature search; determine who you ...

  2. solr 对于 关键字的特殊处理

    public static String transformMetachar(String input){      StringBuffer sb = new StringBuffer();     ...

  3. webpack如何编译ES6打包

    前言:随着ES的普及我们越来越多的开始使用ES6的语法了,当然也随着mvvm框架的流行少不了js模块化,那js模块化又有那些呢 在很早的时候大家都用的命名空间,现在也有人用(库名.类别名.方法名) 后 ...

  4. 2020牛客寒假算法基础集训营1 J. 缪斯的影响力 (矩阵快速幂/费马小定理降幂)

    https://ac.nowcoder.com/acm/problem/200658 f(n) = f(n-1) * f(n-2) * ab ,f的第一项是x,第二项是y. 试着推出第三项是x·y·a ...

  5. flaskapp

    前置知识 https://blog.csdn.net/u013457794/article/details/88997699?depth_1-utm_source=distribute.pc_rele ...

  6. python调用HEG工具批量处理MODIS数据

    下面的代码主要用于使用python语言调用NASA官方的MODIS处理工具HEG进行投影坐标转换与重采样批量处理 主要参考 HEG的用户手册:https://newsroom.gsfc.nasa.go ...

  7. vue学习指南:第十五篇(详细) - Vuex

    Vuex 一.基础 1. Vuex 相当于 vue的数据仓库 2. Vuex 是 vue 的状态管理工具 3. Vuex中的 state 只能通过mutations 改变 4. Vuex很适合做购物车 ...

  8. 高级特征工程II

    以下是Coursera上的How to Win a Data Science Competition: Learn from Top Kagglers课程笔记. Statistics and dist ...

  9. jdk8-》stream⾥的map和filter函数使⽤

    map函数 将流中的每⼀个元素 T(入参) 映射为 R(返回值)(类似类型转换)    类似遍历集合,对集合的每个对象做处理.场景:转换对象,如javaweb开发中集合⾥⾯的DO对象转换为DTO对象 ...

  10. DeepLearningDTU: Building a RNN step by step

    exercise 5: Week 5 - Recurrent Neural Networks Building your Recurrent Neural Network - Step by Step