iOS Accelerate Framework vImage - 性能改进？

Question

我一直在使用 OpenCV 和 Apple 的 Accelerate 框架，发现 Accelerate 的性能很慢，Apple 的文档也很有限。让我们举个例子:

void equalizeHistogram(const cv::Mat &planar8Image, cv::Mat &equalizedImage)
{
    cv::Size size = planar8Image.size();
    vImage_Buffer planarImageBuffer = {
        .width = static_cast<vImagePixelCount>(size.width),
        .height = static_cast<vImagePixelCount>(size.height),
        .rowBytes = planar8Image.step,
        .data = planar8Image.data
    };

    vImage_Buffer equalizedImageBuffer = {
        .width = static_cast<vImagePixelCount>(size.width),
        .height = static_cast<vImagePixelCount>(size.height),
        .rowBytes = equalizedImage.step,
        .data = equalizedImage.data
    };

    TIME_START(VIMAGE_EQUALIZE_HISTOGRAM);
    vImage_Error error = vImageEqualization_Planar8(&planarImageBuffer, &equalizedImageBuffer, kvImageNoFlags);
    TIME_END(VIMAGE_EQUALIZE_HISTOGRAM);
    if (error != kvImageNoError) {
        NSLog(@"%s, vImage error %zd", __PRETTY_FUNCTION__, error);
    }
}

此调用大约需要 20 毫秒。这具有在我的应用程序中无法使用的实际意义。也许直方图的均衡本来就很慢，但我也测试了 BGRA->Grayscale，发现 OpenCV 可以在 ~5ms 内完成，而 vImage 需要 ~20ms。

在测试其他功能时，我发现了一个 project that made a simple slider app用blur function (gist)我清理了测试。大约也是 20 毫秒。

是否有一些技巧可以使这些功能更快？

Answer 1

要使用 equalizeHistogram 函数每秒获得 30 帧，您必须对图像进行去交错处理(从 ARGBxxxx 转换为 PlanarX)并仅均衡 R(ed)G(reen)B(lue)；如果均衡 A(lpha)，帧率将下降到至少 24。

这是完全按照您的要求执行的代码:

- (CVPixelBufferRef)copyRenderedPixelBuffer:(CVPixelBufferRef)pixelBuffer {

CVPixelBufferLockBaseAddress( pixelBuffer, 0 );

unsigned char *base = (unsigned char *)CVPixelBufferGetBaseAddress( pixelBuffer );
size_t width = CVPixelBufferGetWidth( pixelBuffer );
size_t height = CVPixelBufferGetHeight( pixelBuffer );
size_t stride = CVPixelBufferGetBytesPerRow( pixelBuffer );

vImage_Buffer _img = {
    .data = base,
    .height = height,
    .width = width,
    .rowBytes = stride
};

vImage_Error err;
vImage_Buffer _dstA, _dstR, _dstG, _dstB;

err = vImageBuffer_Init( &_dstA, height, width, 8 * sizeof( uint8_t ), kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageBuffer_Init (alpha) error: %ld", err);

err = vImageBuffer_Init( &_dstR, height, width, 8 * sizeof( uint8_t ), kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageBuffer_Init (red) error: %ld", err);

err = vImageBuffer_Init( &_dstG, height, width, 8 * sizeof( uint8_t ), kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageBuffer_Init (green) error: %ld", err);

err = vImageBuffer_Init( &_dstB, height, width, 8 * sizeof( uint8_t ), kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageBuffer_Init (blue) error: %ld", err);

err = vImageConvert_ARGB8888toPlanar8(&_img, &_dstA, &_dstR, &_dstG, &_dstB, kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageConvert_ARGB8888toPlanar8 error: %ld", err);

err = vImageEqualization_Planar8(&_dstR, &_dstR, kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageEqualization_Planar8 (red) error: %ld", err);

err = vImageEqualization_Planar8(&_dstG, &_dstG, kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageEqualization_Planar8 (green) error: %ld", err);

err = vImageEqualization_Planar8(&_dstB, &_dstB, kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageEqualization_Planar8 (blue) error: %ld", err);

err = vImageConvert_Planar8toARGB8888(&_dstA, &_dstR, &_dstG, &_dstB, &_img, kvImageNoFlags);
if (err != kvImageNoError)
    NSLog(@"vImageConvert_Planar8toARGB8888 error: %ld", err);

err = vImageContrastStretch_ARGB8888( &_img, &_img, kvImageNoError );
if (err != kvImageNoError)
    NSLog(@"vImageContrastStretch_ARGB8888 error: %ld", err);

free(_dstA.data);
free(_dstR.data);
free(_dstG.data);
free(_dstB.data);

CVPixelBufferUnlockBaseAddress( pixelBuffer, 0 );

return (CVPixelBufferRef)CFRetain( pixelBuffer );

请注意，我分配了 alpha channel ，尽管我没有对其执行任何操作；这仅仅是因为在 ARGB8888 和 Planar8 之间来回转换需要 alpha channel 缓冲区分配和引用。无论如何，性能和质量都得到了提升。

另请注意，我在将 Planar8 缓冲区转换为单个 ARGB8888 缓冲区后执行对比度拉伸(stretch)；那是因为它比逐个 channel 应用函数更快，就像我对直方图均衡函数所做的那样，并且得到与单独执行相同的结果(对比度拉伸(stretch)函数不会导致与直方图均衡相同的 alpha channel 失真) .