median_filter.py

Go to the documentation of this file.

import ctypes
from typing import Union
 
from ids_peak_icv import lib_loader
from ids_peak_icv.backend.utils import execute_and_map_return_codes
from ids_peak_icv.datatypes.image import Image, _check_and_get_icv_image
from ids_peak_common.datatypes.geometry.size import Size
from ids_peak_ipl import ids_peak_ipl
 
 
class MedianFilter:
    """!
    @brief
    Class for applying a median filter to an image.
 
    The MedianFilter class applies a median filter to reduce noise in an image. It replaces each pixel value with the
    median value of its neighboring pixels, based on a defined kernel size.
 
    @ingroup ids_peak_icv_python_filters
    @since ids_peak_icv 1.1
    """
 
    def __init__(self) -> None:
        """!
        @brief
        Creates an instance of the MedianFilter class with a default kernel size of 3x3.
 
        @since ids_peak_icv 1.1
        """
        self._kernel_size = 3
 
    @property
    def kernel_size(self) -> int:
        """!
        @brief
        Gets the current kernel size used for the median filter.
 
        @return int: The size of the kernel.
 
        @since ids_peak_icv 1.1
        """
        return self._kernel_size
 
    @kernel_size.setter
    def kernel_size(self, kernel_size: int) -> None:
        """!
        @brief
        Sets the size of the kernel used for the median filter.
 
        @param kernel_size: Size of the kernel. Must be an odd number greater than 1.
                            Also, it has to be smaller than the image size.
 
        @since ids_peak_icv 1.1
        """
        self._kernel_size = kernel_size
 
    def process(self, image: Union[Image, ids_peak_ipl.Image]) -> Image:
        """!
        @brief
        Applies the median filter to the given image and returns the filtered result.
 
        This function processes the entire image and reduces noise using the specified kernel size.
        The output image will have the same pixel format and dimensions as the input image.
 
        @note This operation disregards any specified image regions. It processes the entire image.
 
        @param image: The input image to which the median filter will be applied.
                      It can be of type Image or ids_peak_ipl.Image.
        @return Image: The filtered image with reduced noise.
        @exception NotSupportedException:  If the input image is of an unsupported format.
        @exception NotPossibleException:   If the KernelSize is not a valid odd number greater than 1.
        @exception OutOfRangeException:    If the kernel size is bigger than the given image.
 
        @since ids_peak_icv 1.1
        """
        _image = _check_and_get_icv_image(image)
 
        output_image = Image.create_from_pixel_format_and_size(_image._pixel_format,
                                                               Size(_image.width, _image.height))
 
        execute_and_map_return_codes(lib_loader.dll().peak_icv_Filter_Image_Median,
                                     _image._handle,
                                     ctypes.c_size_t(self._kernel_size),
                                     output_image._handle)
 
        return output_image