I can't comment on the top level answer because I don't have enough stackoverflow rep yet, but I just wanted to point out one very critical bug with that implementation.
Some bitmaps can be written with a negative height, so when you try to allocate your image data buffer, your code will crash with std::bad_alloc. Bitmaps with negative height means that the image data is stored top to bottom instead of the traditional bottom to top. Therefore, a slightly better version of the top level answer is (still not including portability for systems with different endianness and size of bytes):
unsigned char* readBMP(char* filename)
{
int i;
FILE* f = fopen(filename, "rb");
unsigned char info[54];
fread(info, sizeof(unsigned char), 54, f); // read the 54-byte header
// extract image height and width from header
int width, height;
memcpy(&width, info + 18, sizeof(int));
memcpy(&height, info + 22, sizeof(int));
int heightSign = 1;
if (height < 0){
heightSign = -1;
}
int size = 3 * width * abs(height);
unsigned char* data = new unsigned char[size]; // allocate 3 bytes per pixel
fread(data, sizeof(unsigned char), size, f); // read the rest of the data at once
fclose(f);
if(heightSign == 1){
for(i = 0; i < size; i += 3)
{
//code to flip the image data here....
}
}
return data;
}