读取BMP文件会返回意外数据

Question

I made a bitmap with the following first row: 0000000000000000000000000000000111111111 (It's a black line on the edge of the bitmap) But when I read the bitmap I get back the following data on the first row: 1000100100001001000010010000100100001001

Actually the first 3 rows contain the 1's and some zeros. The rest of the values are 0's. I use the following code:

#include "stdafx.h"
#include <iostream>
#include <vector>

using namespace std;

int main() {
    FILE* f = fopen("C:\\Users\\Laptop_Chris\\Documents\\kleinObject.bmp", "rb");
    unsigned char info[54];
    fread(info, sizeof(unsigned char), 54, f);

    int size = ((*(int*)&info[18] * 3 + 3) & (~3)) * *(int*)&info[22]; //BMP - HEIGHT: *(int*)&info[22]  -  WIDTH:*(int*)&info[18]

    unsigned char* data = new unsigned char[size];
    fread(data, sizeof(unsigned char), size, f);
    fclose(f);

    vector< vector<bool> > myVector;
    myVector.resize(*(int*)&info[22], vector<bool>(*(int*)&info[18]));

    int i = 0;

    for (auto a = 0; a < *(int*)&info[22]; a++) {
        for (auto q = 0; q < *(int*)&info[18]; q++) {
            if ((int(data[i] & 0xFF) + int(data[i + 1] & 0xFF) + int(data[i + 2] & 0xFF)) > 0) {
                myVector[a][q] = false;
            }
            else {
                myVector[a][q] = true;
            }
            i = i + 3;
            cout << myVector[a][q];
        }
        cout << endl;
    }
    cin.get();
    return 0;
}

Anybody that understands this behavior? Thanks in advance.

Answer 1

Your assumptions are that you have a RGB24 bitmap and that BITMAPINFOHEADER::biHeight > 0 (that is your *(int*)&info[22] ). Note that this means that your image is stored from bottom row to top row.

Furthermore, you handle 4 byte padding when computing size , but you are not handling it when reading out data .

Also note that you are outtputting 1 for R=G=B=0 (ie black) and 0 otherwise. You probably intend the opposite.

You should also add a delete [] data; at the end.

Summary: your code only works for an RGB24 image with biHeight > 0 and (3 * biWidth) % 4 == 0 and you have to be aware that you are outputting from bottom row to top row and 1 for R=G=B=0 and 0 otherwise.