How to edit values of images - c++

Question

I am having some trouble playing around with files. Here is what I am trying to accomplish. I am attempting to filter a PPM image by take the inverse of the blue value (every third value). I can successfully open and write files, but I am experiencing some issues. In my while (myfile.good()) loop, I think only the final 3 numbers get assigned to the variables r, g, and b. What I am trying to do is have every first( out of three) value assigned to the variable r, every second value assigned to g, and every third value assigned to b. However, I want to take 255 - the current b value, and set that as the new b value to apply to filter. Do I have to make 3 separate files (1 for each variable), and then open them all to write them in a 4 file which would serve as the final copy? Or is there a way to copy them all and assign it to a variable? Any help is much appreciated. I am new to c++ so please forgive me. Thank you for your assistance.
Example of values I am attempting to use: http://imgur.com/H6EDFIq
#include <iostream>
#include <cmath>
#include <fstream>
#include <cstdlib>
using namespace std;
int main()
{
char filename[50];
ifstream myfile;
cin.getline(filename, 50);
myfile.open(filename);
if(!myfile.is_open())
{
cout << "File cannot load.";
}
char r [50];
char g [50];
char b [50];
myfile >> r >> g >> b;
while (myfile.good())
{
myfile >> r >> g >> b;
}
myfile.close();
ofstream myfile2;
myfile2.open(filename);
//this is just to test to see what gets written onto the file
//for (int a=0; a<20; a++)
//{
// ** Is this even allowed?? int r = 255 - r;
//myfile2 << r << " " << g << " " << b;
//}
myfile2.close();
return 0;
}

Answer 1

Unless you have a particular need to store the data in memory then it's simpler to just write the data to a new file as you read it.
Here's an example that reads a P3 ppm with 8-bit color entries, inverts the blue channel as you requested, then writes a new file with that data. It doesn't do a ton of error checking but I didn't want to make it any longer than it already is. You'll want to add your own filename prompting and such.
#include <iostream>
#include <string>
#include <fstream>
int main()
{
std::ifstream inFile("lena.ppm");
if(!inFile)
{
std::cerr << "Could not open input file.\n";
return -1;
}
std::string type;
std::string comment;
int width = 0, height = 0, colors = 0;
std::getline(inFile, type);
if(type != "P3")
{
std::cerr << "File is not a P3 format PPM.\n";
return -1;
}
if(inFile.peek() == '#')
{
std::getline(inFile, comment);
}
inFile >> width >> height >> colors;
std::ofstream outFile("lena2.ppm");
if(!outFile)
{
std::cerr << "Could not open output file.\n";
return -1;
}
outFile << type << "\n";
if(!comment.empty())
{
outFile << comment << "\n";
}
outFile << width << " " << height << "\n" << colors << "\n";
for(int y = 0; y < height; ++y)
{
for(int x = 0; x < width; ++x)
{
int r = 0, g = 0, b = 0;
if(!(inFile >> r >> g >> b))
{
std::cerr << "File ended early.\n";
return -1;
}
b = (255 - b);
if(x != 0 && !(x % 5)) //Keep lines under 70 columns per spec.
{
outFile << "\n";
}
outFile << r << " " << g << " " << b << " ";
}
outFile << "\n";
}
return 0;
}

Answer 2

You'll want to check for a .ppm header which is the first line of the .ppm image. Check out .ppm magic numbers, either P3 or P6 and you'll need to check for that. Second line are the dimensions of the image, so you'll need to take that into consideration too.
This is something I worked on earlier so you can get an idea. It might not work immediately so just give it a read.
#include <iostream>
#include <fstream>
#include <sstream>
#include <exception>
int main() {
std::string filename = //whatever your file name is
std::ifstream input(filename.c_str(), std::ios::in | std::ios::binary);
if (input.is_open()) {
std::string line;
std::getline(input, line);
if (line != "P6" || line != "P3" ) { //you're going to want to check if you're using P3 or P6
//print out errors
}
std::stringstream image_dimensions(line);
try {
image_dimensions >> w //your width variable if you want to store it here
image_dimensions >> h //your height variable if you want to store it here
} catch (std::exception &e) {
std::cout << "Format error found in header " << e.what() << std::endl;
return;
}
int size = w*h;
std::getline(input, line);
std::stringstream max_value(line); //max colour value of the image
//you can initialise the vectors here if you want
std::vector<unsigned char> r;
std::vector<unsigned char> g;
std::vector<unsigned char> b;
//we know it max capacity so reserve that size for the vectors
r.reserve(size);
g.reserve(size);
b.reserve(size);
char read_rgb;
for (unsigned int i = 0; i < size; ++i) {
input.read(&read_rgb, 1);
r[i] = (unsigned char) read_rgb;
input.read(&read_rgb, 1);
g[i] = (unsigned char) read_rgb;
input.read(&read_rgb, 1);
b[i] = (unsigned char) read_rgb;
}
}
input.close();
}
You'll want to store r,g,b as arrays of your choosing. Once you've done that, you can just iterate over the array of B and edit it to apply your filter and then just write it to a ppm file.
Also, for error handling, you can always open up the .ppm file using Notepad or Notepad++ and read the image.