Image processing device and method, image capturing device, program, and record medium

An image processing method comprising:

• successively extract a plurality of partial region images, each including a division region constituting a part of a distortion correction target region, from an input image, and successively output the plurality of partial region images or a plurality of images obtained by reducing the plurality of partial region images as reference images each including an image of the division region or a reduced division region corresponding to the division region;
• perform distortion correction on the image of the division region or the reduced division region included in each of the reference images, and successively outputting a plurality of distortion-corrected division region images;
• generate a distortion-corrected image of the distortion correction target region by combining the plurality of distortion-corrected division region images;
• generate a high-frequency component image of the distortion-corrected image;
• calculate a division region scaling ratio regarding each of the plurality of distortion-corrected division region images from a size of the distortion-corrected division region image and a size of the division region or the reduced division region in the reference image corresponding to the distortion-corrected division region image;
• successively specify a plurality of pixels forming the distortion-corrected image as attention pixels;
• calculate a pixel gain regarding the attention pixel from-a first division region scaling ratio regarding a first distortion-corrected division region image including a pixel corresponding to the attention pixel of the distortion-corrected image and second division region scaling ratios regarding one or more distortion-corrected division region images adjoining the first distortion-corrected division region image;
• multiply a pixel value of a pixel of the high-frequency component image corresponding to the attention pixel by the pixel gain of the attention pixel and thereby outputting a pixel value of a pixel of an adjusted high-frequency component image corresponding to the attention pixel;
• and
• output a result of addition of a pixel value of the attention pixel of the distortion-corrected image and the pixel value of the pixel of the adjusted high-frequency component image corresponding to the attention pixel as a pixel value of a pixel of a final distortion-corrected image of the distortion correction target region corresponding to the attention pixel.

A non-transitory computer-readable record medium storing a program which when executed by a processor causes the processor to

• successively extract a plurality of partial region images, each including a division region constituting a part of a distortion correction target region, from an input image, and successively output the plurality of partial region images or a plurality of images obtained by reducing the plurality of partial region images as reference images each including an image of the division region or a reduced division region corresponding to the division region;
• perform distortion correction on the image of the division region or the reduced division region included in each of the reference images, and successively outputting a plurality of distortion-corrected division region images;
• generate a distortion-corrected image of the distortion correction target region by combining the plurality of distortion-corrected division region images;
• generate a high-frequency component image of the distortion-corrected image;
• calculate a division region scaling ratio regarding each of the plurality of distortion-corrected division region images from a size of the distortion-corrected division region image and a size of the division region or the reduced division region in the reference image corresponding to the distortion-corrected division region image;
• successively specify a plurality of pixels forming the distortion-corrected image as attention pixels;
• calculate a pixel gain regarding the attention pixel from a first division region scaling ratio regarding a first distortion-corrected division region image including a pixel corresponding to the attention pixel of the distortion-corrected image and second division region scaling ratios regarding one or more distortion-corrected division region images adjoining the first distortion-corrected division region image;
• multiply a pixel value of a pixel of the high-frequency component image corresponding to the attention pixel by the pixel gain of the attention pixel and thereby outputting a pixel value of a pixel of an adjusted high-frequency component image corresponding to the attention pixel;
• and
• output a result of addition of a pixel value of the attention pixel of the distortion-corrected image and the pixel value of the pixel of the adjusted high-frequency component image corresponding to the attention pixel as a pixel value of a pixel of a final distortion-corrected image of the distortion correction target region corresponding to the attention pixel.