Image processing apparatus, image processing method, and program

An image processing apparatus comprising:

• characteristic value calculation means for calculating a block luminance average value, which is an average value of luminance values in each of spatial blocks obtained by dividing an input image in a space direction, and a block luminance dispersion, which is a dispersion of the luminance values in each of the spatial blocks;
• and
• edge-preserving smoothing means for calculating a general luminance value of pixels of the input image by setting a value obtained by approximately calculating a weighted average value of the luminance values of a noticed block which is the spatial block to which noticed pixels of the input image belong, on the assumption that a distribution of the luminance values of each of the spatial blocks conforms to a Gaussian distribution based on the block luminance average value and the block luminance dispersion of each of the spatial blocks, as the general luminance value of the noticed pixels, the weighted average value being calculated using a weight function, in which a weight of the luminance values increases as the luminance values are closer to the luminance values of the noticed pixels.

An image processing method performed by an image processing apparatus performing an edge-preserving smoothing process on an image, the image processing method comprising the steps of:

• calculating a block luminance average value, which is an average value of luminance values in each of spatial blocks obtained by dividing an input image in a space direction, and a block luminance dispersion, which is a dispersion of the luminance values in each of the spatial blocks;
• and
• calculating a general luminance value of pixels of the input image by setting a value obtained by approximately calculating a weighted average value of the luminance values of a noticed block which is the spatial block to which noticed pixels of the input image belong, on the assumption that a distribution of the luminance values of each of the spatial blocks conforms to a Gaussian distribution based on the block luminance average value and the block luminance dispersion of each of the spatial blocks, as the general luminance value of the noticed pixels, the weighted average value being calculated using a weight function, in which a weight of the luminance values increases as the luminance values are closer to the luminance values of the noticed pixels.

A non-transitory computer-readable storage medium comprising instructions for causing a computer to execute the steps of calculating a block luminance average value, which is an average value of luminance values in each of spatial blocks obtained by dividing an input image in a space direction, and a block luminance dispersion which is a dispersion of the luminance values in each of the spatial blocks;

• and
• calculating a general luminance value of pixels of the input image by setting a value obtained by approximately calculating a weighted average value of the luminance values of a noticed block which is the spatial block to which noticed pixels of the input image belong, on the assumption that a distribution of the luminance values of each of the spatial blocks conforms to a Gaussian distribution based on the block luminance average value and the block luminance dispersion of each of the spatial blocks, as the general luminance value of the noticed pixels, the weighted average value being calculated using a weight function in which a weight of the luminance values increases as the luminance values are closer to the luminance values of the noticed pixels.

An image processing apparatus comprising:

• a characteristic value calculation unit calculating a block luminance average value, which is an average value of luminance values in each of spatial blocks obtained by dividing an input image in a space direction, and a block luminance dispersion, which is a dispersion of the luminance values in each of the spatial blocks;
• and
• an edge-preserving smoothing unit calculating a general luminance value of pixels of the input image by setting a value obtained by approximately calculating a weighted average value of the luminance values of a noticed block which is the spatial block to which noticed pixels of the input image belong, on the assumption that a distribution of the luminance values of each of the spatial blocks conforms to a Gaussian distribution based on the block luminance average value and the block luminance dispersion of each of the spatial blocks, as the general luminance value of the noticed pixels, the weighted average value being calculated using a weight function, in which a weight of the luminance values increases as the luminance values are closer to the luminance values of the noticed pixels.