如何用C++程序打开bmp文件

[猪皮冻]

如何用C++写程序打开一幅BMP格式的图片啊，最好是二值的，八位的，和真彩色的程序，给提提思路也好，刚接触，一点思路都没有，各位帮帮忙吧

[luciferisnotsatan]

fopen，二进制打开文件

[vilnies]

LZ的问题应该不是这么简单吧。BMP文件有格式啊，可以根据格式读数据就行,可以看看GDI编程的有关书籍，都有例子的

[cbib_cat]

可以用OpenCV处理图像，很方便的。

[iblold]

// pdat是RGBA格式,w图像的宽,h高
void Bmp2RGBA(char* caFileName, DWORD** pdat, int& w, int& h)
{
	FILE* pf = fopen(caFileName, "rb");
	if (pf == NULL) 
		return;

	BITMAPFILEHEADER bm;
	fread(&bm, 1, sizeof(BITMAPFILEHEADER), pf);

	if (bm.bfType!=*(WORD*)"BM")
	{	
		fclose(pf);
		return;
	}
	BITMAPINFOHEADER bmif;
	fread(&bmif, 1, sizeof(BITMAPINFOHEADER), pf);
	int bmx = bmif.biWidth;
	int bmy = bmif.biHeight>0?bmif.biHeight:(-bmif.biHeight);

	if (bmif.biBitCount < 8 || bmif.biCompression != 0) 
	{
		fclose(pf);
		return;
	}

	RGBQUAD *pRgb = NULL;
	if (bmif.biBitCount == 8) // 256色bmp
	{
		pRgb = (RGBQUAD*) new BYTE[bmif.biClrUsed * sizeof(RGBQUAD)];
		fread(pRgb, bmif.biClrUsed * sizeof(RGBQUAD), 1, pf);
	}

	BOOL bAlpha = FALSE;
	if (bmif.biBitCount == 32)	bAlpha = TRUE;// 带Alpha通道

	DWORD dwRealWidth = (((bmx*bmif.biBitCount)+31)>>5)<<2;
	int iBitSteps = bmif.biBitCount>>3;
	DWORD iSize = dwRealWidth*bmy;

	BYTE* pBmpData = NULL;
	int readlines = bmy;
	int readszie = iSize;
	int readtimes = bmy/readlines;
	if (iSize > 128*1024*1024)
	{
		readlines = (128*1024*1024/dwRealWidth);
		readszie = readlines*dwRealWidth;
		readtimes = bmy/readlines + 1;
	}


	pBmpData = new BYTE[readszie];
	fseek(pf, bm.bfOffBits, SEEK_SET);
	fread(pBmpData, 1, readszie, pf);

	w = bmx;
	h = bmy;
	*pdat = new DWORD[w * h];

	int iy,ix;
	BYTE btPil;
	char caTmp[4];
	DWORD dwDataOff = 0;
	int picy = 0;
	for(int n = 0; n < readtimes; ++n)
	{
		for (iy = 0; iy < readlines; ++iy)
		{
			for (ix = 0; ix < bmx; ++ix)
			{
				if (pRgb == NULL)
				{
					caTmp[0] = pBmpData[(iy * dwRealWidth + ix * iBitSteps) + 2];
					caTmp[1] = pBmpData[(iy * dwRealWidth + ix * iBitSteps) + 1];
					caTmp[2] = pBmpData[(iy * dwRealWidth + ix * iBitSteps)];
				}
				else
				{
					btPil = pBmpData[(iy * dwRealWidth + ix * iBitSteps)];
					if (bmif.biClrUsed != 0)
					{
						caTmp[0] = pRgb[btPil].rgbRed;
						caTmp[1] = pRgb[btPil].rgbGreen;
						caTmp[2] = pRgb[btPil].rgbBlue;
					}
					else
					{
						caTmp[0] = caTmp[1] = caTmp[2] = btPil;
					}

				}

				if (bmif.biHeight < 0)		// 有时bmp图像的高会是负值，这时则不用颠倒 
				{
					dwDataOff = (iy + picy) * bmx + ix;
				}
				else
				{
					dwDataOff = (bmy - (iy + picy) - 1) * bmx + ix;
				}

				caTmp[3] = 255;

				if (bAlpha)
				{
					caTmp[3] = pBmpData[(iy * dwRealWidth + ix * iBitSteps) + 3];
				}

				*pdat[dwDataOff] = *(DWORD*)caTmp;

			}
		}
		picy += readlines;
		int realbits = fread(pBmpData, 1, readszie, pf);
		if (realbits < readszie) readlines = realbits / dwRealWidth;
	}

	if (pf)
		fclose(pf);
	pf = NULL;
}

可以读取8、24、32位bmp,

2位或者256级灰度的图可以在上面函数中添加只要读的时候改个步长就行了

[iblold]

最后加上两行

delete pBmpData;
delete pRgb;
刚才整理的时候误删了