android camera之nv21旋转

这周做的一个android的camera开发，需要获取到视频帧数据，并且需要是nv21格式的byte数组，并且视频帧的图像需要是正方向的。和android相机打过交道的都清楚，android的camera获取到的图片都是横向的，因此，需要进行旋转，对于图像的旋转，其实bitmap这个类已经可以帮我们实现了，但是前提是你需要将你的数据格式转换为Bitmap才行，但是我们如果通过setPreviewCallback来获取视频帧，获取到的图片都是nv21，如果装换为bitmap后，又很难的转换为nv21格式的数据。因此则需要面临一个问题，如歌旋转nv21格式的byte数组，首先先来讲下nv21格式，讲到nv21就也要说说yuv240和nv12：

NV12、NV21（属于YUV420）

aaarticlea/png;base64,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" alt="" />

NV12和NV21属于YUV420格式，是一种two-plane模式，即Y和UV分为两个Plane，但是UV（CbCr）为交错存储，而不是分为三个plane。其提取方式与上一种类似，即Y'00、Y'01、Y'10、Y'11共用Cr00、Cb00

YUV420 planar数据存储， 以720×488大小图象YUV420 planar为例，

其存储格式是： 共大小为(720×480×3>>1)字节，

分为三个部分: Y分量：　　    　(720×480)个字节   U(Cb)分量：　　(720×480>>2)个字节　　　　 V(Cr)分量：　　 (720×480>>2)个字节

三个部分内部均是行优先存储，三个部分之间是Y,U,V 顺序存储。

即YUV数据的0－－720×480字节是Y分量值，    720×480－－720×480×5/4字节是U分量    720×480×5/4 －－720×480×3/2字节是V分量。

4 ：2： 2 和4：2：0 转换：

最简单的方式：

YUV4:2:2 ---> YUV4:2:0  Y不变，将U和V信号值在行(垂直方向)在进行一次隔行抽样。 YUV4:2:0 ---> YUV4:2:2  Y不变，将U和V信号值的每一行分别拷贝一份形成连续两行数据。

在YUV420中，一个像素点对应一个Y，一个4X4的小方块对应一个U和V。对于所有 YUV420图像，它们的Y值排列是完全相同的，因为只有Y的图像就是灰度图像。YUV420sp与YUV420p的数据格式它们的UV排列在原理上是完 全不同的。420p它是先把U存放完后，再存放V，也就是说UV它们是连续的。而420sp它是UV、UV这样交替存放的。(见下图) 有了上面的理论，我就可以准确的计算出一个YUV420在内存中存放的大小。 width * hight =Y（总和） U = Y / 4   V = Y / 4

所以YUV420 数据在内存中的长度是 width * hight * 3 / 2，

假设一个分辨率为8X4的YUV图像，它们的格式如下图：

图：YUV420sp格式

图：YUV420p数据格式如下图

具体的旋转代码如下：

public static byte[] rotateYUV420Degree90(byte[] data, int imageWidth, int imageHeight) {
    byte[] yuv = new byte[imageWidth * imageHeight * 3 / 2];
    int i = 0;
    for (int x = 0; x < imageWidth; x++) {
        for (int y = imageHeight - 1; y >= 0; y--) {
            yuv[i] = data[y * imageWidth + x];
            i++;
        }
    }
    i = imageWidth * imageHeight * 3 / 2 - 1;
    for (int x = imageWidth - 1; x > 0; x = x - 2) {
        for (int y = 0; y < imageHeight / 2; y++) {
            yuv[i] = data[(imageWidth * imageHeight) + (y * imageWidth) + x];
            i--;
            yuv[i] = data[(imageWidth * imageHeight) + (y * imageWidth)
                    + (x - 1)];
            i--;
        }
    }
    return yuv;
}

private static byte[] rotateYUV420Degree180(byte[] data, int imageWidth, int imageHeight) {
    byte[] yuv = new byte[imageWidth * imageHeight * 3 / 2];
    int i = 0;
    int count = 0;
    for (i = imageWidth * imageHeight - 1; i >= 0; i--) {
        yuv[count] = data[i];
        count++;
    }
    i = imageWidth * imageHeight * 3 / 2 - 1;
    for (i = imageWidth * imageHeight * 3 / 2 - 1; i >= imageWidth
            * imageHeight; i -= 2) {
        yuv[count++] = data[i - 1];
        yuv[count++] = data[i];
    }
    return yuv;
}

public static byte[] rotateYUV420Degree270(byte[] data, int imageWidth,
                                     int imageHeight) {
    byte[] yuv = new byte[imageWidth * imageHeight * 3 / 2];
    int nWidth = 0, nHeight = 0;
    int wh = 0;
    int uvHeight = 0;
    if (imageWidth != nWidth || imageHeight != nHeight) {
        nWidth = imageWidth;
        nHeight = imageHeight;
        wh = imageWidth * imageHeight;
        uvHeight = imageHeight >> 1;// uvHeight = height / 2
    }

    int k = 0;
    for (int i = 0; i < imageWidth; i++) {
        int nPos = 0;
        for (int j = 0; j < imageHeight; j++) {
            yuv[k] = data[nPos + i];
            k++;
            nPos += imageWidth;
        }
    }
    for (int i = 0; i < imageWidth; i += 2) {
        int nPos = wh;
        for (int j = 0; j < uvHeight; j++) {
            yuv[k] = data[nPos + i];
            yuv[k + 1] = data[nPos + i + 1];
            k += 2;
            nPos += imageWidth;
        }
    }
    return rotateYuv420Degree180(rotateYuv420Degree90(data, imageWidth, imageHeight), imageWidth, imageHeight);
}

然后，如果你想要查看旋转后的图像，则通过以下代码即可：

YuvImage yuvimage = new YuvImage(
                data,
                ImageFormat.NV21,
                width,
                height,
                null);
        baos = new ByteArrayOutputStream();
        yuvimage.compressToJpeg(new Rect(0, 0, width, height), 100, baos);// 80--JPG图片的质量[0-100],100最高
        rawImage = baos.toByteArray();
        //将rawImage转换成bitmap
        BitmapFactory.Options options = new BitmapFactory.Options();
        options.inPreferredConfig = Bitmap.Config.RGB_565;
        bitmap = BitmapFactory.decodeByteArray(rawImage, 0, rawImage.length, options);

在这里边我们需要注意的是我们的宽和高，需要用转移后的，不可使用转移前的，否则会出现看到的图片有重影的现象。