Image processing apparatus, method, and program

Abstract

An image processing apparatus, including a region setting unit for setting regions on a processing target image in which each pixel of the image is included according to a level of each pixel, a low frequency image generation unit for generating a low frequency image of the processing target image, a gain calculation unit for calculating a gain for each pixel of the processing target image such that the lower the level the greater the gain, wherein the unit calculates the gain such that pixels in each of the regions in which each pixel of the processing target image is included have substantially the same gain based on a region setting result and a level of each pixel of the low frequency image, and a processing unit for generating a processed image by performing dynamic range compression on the processing target image based on the gain.

Description

BACKGROUND OF THE INVENTION[0001]1. Field of the Invention[0002]The present invention relates to an image processing method and apparatus for compressing the dynamic range of an image by changing the gain of the image. The invention also relates to a computer readable recording medium on which is recorded a program for causing a computer to perform the image processing method.[0003]2. Description of the Related Art[0004]Reproduction of images after performing appropriate image processing is common practice in various fields. For example, a dynamic range compression method is proposed. The method generates a low frequency image of an image and reduces the contrast of a high luminance region, a low luminance region, or the entire image such that the difference between the maximum luminance value and minimum luminance value, i.e., dynamic range is reduced using the low frequency image. More specifically, a low frequency image of an image is generated, then a gain is calculated from the...

AI Technical Summary

Benefits of technology

[0040]In this way, according to the first image processing apparatus and method, substantially the same gain is calculated for pixels in each of the one or more regions of the processing target image in which each pixel of the processing target image is included. Consequently, dynamic range compression of the processing target image does not cause overshooting and undershooting arising from a large pixel value variation adjacent to the boundary of the one or more regions in which each pixel of the image is included. Therefore, a high quality processed image without overshooting and undershooting may be obtained.

[0043]Here, the first and second low frequency images are close to each other in frequency band, so that the difference in blur degree is small. Therefore, the pixel level variation range in the first low frequency image becomes small at the boundary between the regions in which each pixel of the second low frequency image is included, so that the gains for the second low frequency image may be calculated appropriately.

[0045]Consequently, even where the frequency band of the low frequency image is low, by gradually increasing the gain to be calculated and frequency band for region splitting, the gain according to a global brightness variation of the processing target image may be calculated without being influenced by the variation range of pixel levels at the boundary between the regions in which each pixel of the processing target image is included. Accordingly, a high quality processed image without overshooting and undershooting may be obtained.

Problems solved by technology

In the dynamic range compression method described above, however, artifacts, such as overshooting, undershooting, and the like occur adjacent to an edge on an image, which cause a problem of image quality degradation.

The method described in Patent Document 1, however, is not able to completely eliminate overshooting or undershooting adjacent to an edge included in an image, since the dynamic range compression is performed based on frequency separation of the image.

Images