# separableConvolve(horizontalKernel:verticalKernel:bias:edgeMode:destination:) Performs separable convolution on a 32-bit planar pixel buffer. ``` func separableConvolve(horizontalKernel: [Float], verticalKernel: [Float], bias: Float = 0, edgeMode: vImage.EdgeMode, destination: vImage.PixelBuffer) ``` ## Parameters `horizontalKernel` The 1D horizontal convolution kernel. `verticalKernel` The 1D vertical convolution kernel. `bias` A value that the operation adds to each element in the convolution result, before performing any clipping. `edgeMode` The convolution edge mode. The background color must be a single [`Pixel_F`](/documentation/Accelerate/Pixel_F) value. `destination` The destination pixel buffer. ## Discussion The following code shows how to apply a Gaussian blur using separable convolution to a planar buffer: ```swift let srcImage = imageLiteral(resourceName: " ... ").cgImage( forProposedRect: nil, context: nil, hints: nil)! var cgImageFormat = vImage_CGImageFormat( bitsPerComponent: 32, bitsPerPixel: 32 * 1, colorSpace: CGColorSpaceCreateDeviceGray(), bitmapInfo: CGBitmapInfo(rawValue: kCGBitmapByteOrder32Host.rawValue | CGBitmapInfo.floatComponents.rawValue | CGImageAlphaInfo.none.rawValue))! let src = try vImage.PixelBuffer( cgImage: srcImage, cgImageFormat: &cgImageFormat, pixelFormat: vImage.PlanarF.self) let dest = vImage.PixelBuffer( size: src.size, pixelFormat: vImage.PlanarF.self) src.separableConvolve( horizontalKernel: vImage.ConvolutionKernel.gaussian1Dx7, verticalKernel: vImage.ConvolutionKernel.gaussian1Dx7, edgeMode: .truncateKernel, destination: dest) let outputImage = dest.makeCGImage(cgImageFormat: cgImageFormat) ``` ## See Also --- Copyright © 2026 Apple Inc. All rights reserved. | [Terms of Use](https://www.apple.com/legal/internet-services/terms/site.html) | [Privacy Policy](https://www.apple.com/privacy/privacy-policy)