Image processing apparatus, image processing method, and program
View Patent ↗Provided is an image processing apparatus including an infinite impulse response (IIR) mean value calculation section for calculating a mean value of signal values of reference pixels around a correction target pixel according to an IIR filter application process, an IIR variance value calculation section for calculating a variance value of the signal values of the reference pixels around the correction target pixel according to the IIR filter application process, an edge-preserving smoothing processing section for receiving the mean and variance values of the reference pixels and executing an edge-preserving smoothing process to which the mean and variance values are applied, and an IIR filter coefficient calculation section for updating an IIR filter coefficient to be applied to the IIR mean value calculation section and the IIR variance value calculation section according to a signal value of a pixel constituting an image.
1. An image processing apparatus comprising:
an infinite impulse response (IIR) mean value calculation section for calculating a mean value of signal values of reference pixels around a correction target pixel according to an IIR filter application process;
an IIR variance value calculation section for calculating a variance value of the signal values of the reference pixels around the correction target pixel according to the IIR filter application process;
an edge-preserving smoothing processing section for receiving the mean and variance values of the reference pixels and executing an edge-preserving smoothing process to which the mean and variance values are applied; and
an IIR filter coefficient calculation section for updating an IIR filter coefficient to be applied to the IIR mean value calculation section and the IIR variance value calculation section according to a signal value of a pixel constituting an image.
2. The image processing apparatus according to claim 1 , wherein the edge-preserving smoothing processing section executes the edge-preserving smoothing process based on a Wiener filter application process.
3. The image processing apparatus according to claim 1 , wherein the edge-preserving smoothing processing section calculates a signal value after correction of the correction target pixel by a process of blending the mean value generated by the IIR mean value calculation section and a signal value of the correction target pixel according to the variance value calculated by the IIR variance value calculation section.
4. The image processing apparatus according to claim 1 , wherein the edge-preserving smoothing processing section is configured to calculate a signal value after correction of the correction target pixel by a process of blending the mean value generated by the IIR mean value calculation section and a signal value of the correction target pixel, and executes the blend process in which a blend rate is set to (v1−v2)/v2 when the variance value calculated by the IIR variance value calculation section is v1 and a parameter by which strength of edge-preserving smoothing is controlled is v2.
5. The image processing apparatus according to claim 1 , wherein the edge-preserving smoothing processing section is configured to calculate a signal value after correction of the correction target pixel by a process of blending the mean value generated by the IIR mean value calculation section and a signal value of the correction target pixel, and executes the blend process in which a blend rate is set to v1/(v1+v2) when the variance value calculated by the IIR variance value calculation section is v1 and a parameter by which strength of edge-preserving smoothing is controlled is v2.
6. The image processing apparatus according to claim 1 , wherein the IIR filter coefficient calculation section sequentially calculates a new IIR filter coefficient according to an input signal value.
7. The image processing apparatus according to claim 1 , wherein the IIR filter coefficient calculation section sequentially calculates a new IIR filter coefficient according to the variance value calculated by the IIR variance value calculation section.
8. The image processing apparatus according to claim 1 , wherein the IIR filter coefficient calculation section sequentially calculates a new IIR filter coefficient based on a difference between a signal value of the correction target pixel and the mean value calculated by the IIR mean value calculation section.
9. The image processing apparatus according to claim 1 , wherein the IIR filter coefficient calculation section sequentially calculates a new IIR filter coefficient based on a difference between a signal value of the correction target pixel and a signal value of a previous input pixel.
10. The image processing apparatus according to claim 1 , wherein:
a two-dimensional image signal is input as a processing target image, and
the edge-preserving smoothing process is applied in at least one of horizontal and vertical directions.
11. The image processing apparatus according to claim 1 , wherein:
a two-dimensional image signal is input as a processing target image,
the edge-preserving smoothing process is applied in at least one of horizontal and vertical directions, and
an isotropic edge-preserving smoothing process is executed in the other direction.
12. The image processing apparatus according to claim 1 , wherein:
a luminance signal image and a color signal image are separately input as a processing target image, and
a process in which strength of edge-preserving smoothing is changed for each image is executed.
13. An image processing method to be executed in an image processing apparatus, comprising:
calculating, by an IIR mean value calculation section, a mean value of signal values of reference pixels around a correction target pixel according to an IIR filter application process;
calculating, by an IIR variance value calculation section, a variance value of the signal values of the reference pixels around the correction target pixel according to the IIR filter application process;
receiving, by an edge-preserving smoothing processing section, the mean and variance values of the reference pixels and executing an edge-preserving smoothing process to which the mean and variance values are applied; and
updating, by an IIR filter coefficient calculation section, an IIR filter coefficient to be applied in the IIR mean value calculation step and the IIR variance value calculation step according to a signal value of a pixel constituting an image.
14. At least one non-transitory computer-readable storage having encoded thereon executable instructions that, when executed by at least one computer, cause the at least one computer to carry out a method for causing an image processing apparatus to execute image processing, the method comprising:
calculating, by an IIR mean value calculation section, a mean value of signal values of reference pixels around a correction target pixel according to an IIR filter application process;
calculating, by an IIR variance value calculation section, a variance value of the signal values of the reference pixels around the correction target pixel according to the IIR filter application process;
receiving, by an edge-preserving smoothing processing section, the mean and variance values of the reference pixels and executing an edge-preserving smoothing process to which the mean and variance values are applied; and
updating, by an IIR filter coefficient calculation section, an IIR filter coefficient to be applied in the IIR mean value calculation step and the IIR variance value calculation step according to a signal value of a pixel constituting an image.