Example: Filter an image¶

fromPILimportImageFilterim1=im.filter(ImageFilter.BLUR)im2=im.filter(ImageFilter.MinFilter(3))im3=im.filter(ImageFilter.MinFilter)# same as MinFilter(3)

Filters¶

Pillow provides the following set of predefined image enhancement filters:

• BLUR

• CONTOUR

• DETAIL

• EDGE_ENHANCE

• EDGE_ENHANCE_MORE

• EMBOSS

• FIND_EDGES

• SHARPEN

• SMOOTH

• SMOOTH_MORE

classPIL.ImageFilter.Color3DLUT(size:int|tuple[int,int,int],table:Sequence[float]|Sequence[Sequence[int]]|NumpyArray,channels:int=3,target_mode:str|None=None,**kwargs:bool)[source]¶

Three-dimensional color lookup table.

Transforms 3-channel pixels using the values of the channels as coordinatesin the 3D lookup table and interpolating the nearest elements.

This method allows you to apply almost any color transformationin constant time by using pre-calculated decimated tables.

Added in version 5.2.0.

Parameters:

• size – Size of the table. One int or tuple of (int, int, int).Minimal size in any dimension is 2, maximum is 65.

• table – Flat lookup table. A list ofchannels*size**3float elements or a list ofsize**3 channels-sizedtuples with floats. Channels are changed first,then first dimension, then second, then third.Value 0.0 corresponds lowest value of output, 1.0 highest.

• channels – Number of channels in the table. Could be 3 or 4.Default is 3.

• target_mode – A mode for the result image. Should have not lessthanchannels channels. Default isNone,which means that mode wouldn’t be changed.

classmethodgenerate(size:int|tuple[int,int,int],callback:Callable[[float,float,float],tuple[float,...]],channels:int=3,target_mode:str|None=None)→Color3DLUT[source]¶

Generates new LUT using provided callback.

Parameters:

• size – Size of the table. Passed to the constructor.

• callback – Function with three parameters which correspondthree color channels. Will be calledsize**3times with values from 0.0 to 1.0 and should returna tuple withchannels elements.

• channels – The number of channels which should return callback.

• target_mode – Passed to the constructor of the resultinglookup table.

transform(callback:Callable[...,tuple[float,...]],with_normals:bool=False,channels:int|None=None,target_mode:str|None=None)→Color3DLUT[source]¶

Transforms the table values using provided callback and returnsa new LUT with altered values.

Parameters:

• callback – A function which takes old lookup table valuesand returns a new set of values. The numberof arguments which function should take isself.channels or3+self.channelsifwith_normals flag is set.Should return a tuple ofself.channels orchannels elements if it is set.

• with_normals – If true,callback will be called withcoordinates in the color cube as the firstthree arguments. Otherwise,callbackwill be called only with actual color values.

• channels – The number of channels in the resulting lookup table.

• target_mode – Passed to the constructor of the resultinglookup table.

classPIL.ImageFilter.BoxBlur(radius:float|Sequence[float])[source]¶

Blurs the image by setting each pixel to the average value of the pixelsin a square box extending radius pixels in each direction.Supports float radius of arbitrary size. Uses an optimized implementationwhich runs in linear time relative to the size of the imagefor any radius value.

Parameters:

radius –

Size of the box in a direction. Either a sequence of two numbers forx and y, or a single number for both.

Radius 0 does not blur, returns an identical image.Radius 1 takes 1 pixel in each direction, i.e. 9 pixels in total.

classPIL.ImageFilter.GaussianBlur(radius:float|Sequence[float]=2)[source]¶

Blurs the image with a sequence of extended box filters, whichapproximates a Gaussian kernel. For details on accuracy see<https://www.mia.uni-saarland.de/Publications/gwosdek-ssvm11.pdf>

Parameters:

radius – Standard deviation of the Gaussian kernel. Either a sequence of twonumbers for x and y, or a single number for both.

classPIL.ImageFilter.UnsharpMask(radius:float=2,percent:int=150,threshold:int=3)[source]¶

Unsharp mask filter.

See Wikipedia’s entry ondigital unsharp masking for an explanation ofthe parameters.

Parameters:

• radius – Blur Radius

• percent – Unsharp strength, in percent

• threshold – Threshold controls the minimum brightness change thatwill be sharpened

classPIL.ImageFilter.Kernel(size:tuple[int,int],kernel:Sequence[float],scale:float|None=None,offset:float=0)[source]¶

Create a convolution kernel. This only supports 3x3 and 5x5 integer and floatingpoint kernels.

Kernels can only be applied to “L” and “RGB” images.

Parameters:

• size – Kernel size, given as (width, height). This must be (3,3) or (5,5).

• kernel – A sequence containing kernel weights. The kernel will be flippedvertically before being applied to the image.

• scale – Scale factor. If given, the result for each pixel is divided by thisvalue. The default is the sum of the kernel weights.

• offset – Offset. If given, this value is added to the result, after it hasbeen divided by the scale factor.

classPIL.ImageFilter.RankFilter(size:int,rank:int)[source]¶

Create a rank filter. The rank filter sorts all pixels ina window of the given size, and returns therank’th value.

Parameters:

• size – The kernel size, in pixels.

• rank – What pixel value to pick. Use 0 for a min filter,size*size/2 for a median filter,size*size-1for a max filter, etc.

classPIL.ImageFilter.MedianFilter(size:int=3)[source]¶

Create a median filter. Picks the median pixel value in a window with thegiven size.

Parameters:

size – The kernel size, in pixels.

classPIL.ImageFilter.MinFilter(size:int=3)[source]¶

Create a min filter. Picks the lowest pixel value in a window with thegiven size.

Parameters:

size – The kernel size, in pixels.

classPIL.ImageFilter.MaxFilter(size:int=3)[source]¶

Create a max filter. Picks the largest pixel value in a window with thegiven size.

Parameters:

size – The kernel size, in pixels.

classPIL.ImageFilter.ModeFilter(size:int=3)[source]¶

Create a mode filter. Picks the most frequent pixel value in a box with thegiven size. Pixel values that occur only once or twice are ignored; if nopixel value occurs more than twice, the original pixel value is preserved.

Parameters:

size – The kernel size, in pixels.

classPIL.ImageFilter.Filter[source]¶

An abstract mixin used for filtering images(for use withfilter()).

Implementors must provide the following method:

filter(self,image)[source]¶

Applies a filter to a single-band image, or a single band of an image.

Returns:

A filtered copy of the image.

classPIL.ImageFilter.MultibandFilter[source]¶

An abstract mixin used for filtering multi-band images(for use withfilter()).

Implementors must provide the following method:

filter(self,image)¶

Applies a filter to a multi-band image.

Returns:

A filtered copy of the image.