[Opencv] pixel-operated digital image processing and opencv Digital Image Processing

[Opencv] pixel-operated digital image processing and opencv Digital Image Processing

A few days ago, the matlab was used to process digital images and matrix operations. If you forget linear algebra, it would actually be GG.

After using matlab, I decided to use opencv. C ++ seems to be a good tool for processing digital images.

1. Install opencv on ubuntu

The tutorials are detailed on Baidu.

2. pixel processing image:

(1) based on the input data, use the adjacent sampling interpolation method to scale the image

Principle of interpolation method of adjacent sampling: http://blog.chinaunix.net/uid-27675161-id-3452025.html

(2) Change the gray-scale resolution of Digital Images Based on Input

(3) This experiment is implemented using grayscale images.

3. paste the code and comments directly.

1 // my_hw1.cpp 2 # include <stdio. h> 3 # include <iostream> 4 # include <opencv2/opencv. hpp> 5 # include <opencv2/core. hpp> 6 # include <opencv2/highgui. hpp> 7 8 using namespace cv; 9 using namespace std; 10 11 string my_pic = "07.png"; 12 Mat image = imread (my_pic, CV_LOAD_IMAGE_GRAYSCALE); 13 14 void on_trackbar (int p, void *) {15 16 ********** * **************** 17 // The image is reloaded every time the callback is performed. 18 Mat src = imread (my_pic, CV_LOAD_IMAGE_GRAYSCALE); 19 20 // get the value of three slide bars 21 int c = cvGetTrackbarPos ("Width", "after change"); 22 int r = cvGetTrackbarPos ("Height ", "after change"); 23 int level = cvGetTrackbarPos ("Level", "after change"); 24 25 // record the width and height of the original graph, and the corresponding vertex sx, sy (used in the neighboring sampling interpolation method) 26 int sh = src. rows; 27 int sw = src. cols; 28 int sx, sy; 29 30 // reset the new image 31 Mat dst (r, c, src. type (); 32 33 // fill in the new image's pixel 34 for (int I = 0; I <r; I ++) {35 for (int j = 0; j <c; j ++) {36 double x = I/(r + 0.0); 37 double y = j/(c + 0.0); 38 sx = sh * x; 39 sy = sw * y; 40 dst. at <uchar> (I, j) = src. at <uchar> (sx, sy ); 41} 42} 43 44 // ********************** Quantization *********** * ************ 45 46 // make the level value legal 47 if (level <= 256 & level> = 128) level = 128; 48 else if (level <128 & level> = 64) level = 64; 49 else if (level <64 & level> = 32) level = 32; 50 else if (level <32 & level> = 16) level = 16; 51 else if (level <16 & level> = 8) level = 8; 52 else if (level <8 & level> = 4) level = 4; 53 else level = 2; 54 55 int channels = dst. channels (); 56 int nrows = dst. rows; 57 int ncols = dst. cols * channels; 58 59 uchar table [256]; 60 int degree = 255/(level-1); 61 int number = 256/level; 62 int count = 0; 63 int value = 0; 64 65 // set the required grayscale ing 66 for (int I = 0; I <256; I ++, count ++) {67 if (count <number) table [I] = value; 68 else {69 count = 0; 70 value + = degree; 71 table [I] = value; 72} 73} 74 75 if (src. isContinuous () {76 ncols * = nrows; 77 nrows = 1; 78} 79 80 for (int I = 0; I <nrows; I ++) {81 uchar * p = dst. ptr <uchar> (I); 82 for (int j = 0; j <ncols; j ++) {83 p [j] = table [p [j]; 84} 85} 86 87 // cout <level <"" <c <"<r <" \ n "; 88 imshow ("after change", dst); 89} 90 91 int main (int argc, char ** argv) {92 imshow ("before change", image ); 93 94 int width = 384; 95 int height = 256; 96 int level = 256; 97 namedWindow ("after change", CV_WINDOW_AUTOSIZE); 98 createTrackbar ("Width ", "after change", & width, 800, on_trackbar); 99 createTrackbar ("Height", "after change", & height, 500, on_trackbar); 100 createTrackbar ("Level ", "after change", & level, 256, on_trackbar); // slider101 102 on_trackbar (0, 0); 103 104 waitKey (); 105 return 0; 106} 107

In addition, the compilation method I selected is cmake + make;

CMakeLists.txt

1 release (VERSION 2.8) 2 project (my_hw1) 3 find_package (OpenCV REQUIRED) 4 add_executable (my_hw1 my_hw1.cpp) 5 target_link_libraries (my_hw1 $ {OpenCV_LIBS })CMakeLists.txt

Place the corresponding image in the same directory, enter the folder, and enter the command line

Cmake.

Make

./My_hw1

You can execute it ~~

 

Finally, the following figure shows the running result: