CV-0.3.5.4: OpenCV based machine vision library

Safe HaskellSafe-Infered

CV.Filters

Description

This module is a collection of various image filters

Synopsis

Documentation

gaussian, blurNS, blur :: (Int, Int) -> Image GrayScale D32 -> Image GrayScale D32Source

Create a new image by applying gaussian, or uniform smoothing.

gaussianOp, blurOp :: (Int, Int) -> ImageOperation GrayScale D32Source

Image operation which applies gaussian or unifarm smoothing with a given window size to the image.

bilateral :: (Int, Int) -> (Int, Int) -> Image a D8 -> Image a D8Source

Apply bilateral filtering

getCentralMoment :: CInt -> (CInt, CInt) -> Image c d -> Image channels depthSource

getAbsCentralMoment :: CInt -> (CInt, CInt) -> Image c d -> Image channels depthSource

getMoment :: CInt -> (CInt, CInt) -> Image c d -> Image channels depthSource

selectiveAvg :: (Int, Int) -> Double -> Image GrayScale D32 -> Image GrayScale D32Source

A selective average filter is an edge preserving noise reduction filter. It is a standard gaussian filter which ignores pixel values that are more than a given threshold away from the filtered pixel value.

convolve2D :: (Point2D anchor, ELP anchor ~ Int) => Matrix D32 -> anchor -> Image GrayScale D32 -> Image GrayScale D32Source

convolve2DI :: (CInt, CInt) -> Image c1 d1 -> Image c d -> Image channels depthSource

haar :: IntegralImage -> (Int, Int, Int, Int) -> Image GrayScale D32Source

Filter the image with box shaped averaging mask.

haarAt :: IntegralImage -> (Int, Int, Int, Int) -> DoubleSource

Get an average of a given region.

data IntegralImage Source

 A type for storing integral images. Integral image stores for every pixel the sum of pixels above and left of it. Such images are used for significantly accelerating the calculation of area averages.

integralImage :: Image GrayScale D32 -> IntegralImageSource

 Calculate the integral image from the given image.

verticalAverage :: Image GrayScale D32 -> Image GrayScale D32Source

Replace pixel values by the average of the row.