Unsharp masking based on the multiscale gradient planes

Abstract

SUMMARYIn this paper, control of emphasis by the multiscale brightness gradient is introduced for unsharp masking. This provides an improved image emphasis method for realizing effective emphasis for any image with superposed Gaussian noise and low-contrast images. Unsharp masking is a method of obtaining an enhanced image by applying a high-pass filter to blurred images so that the obtained high-pass components are superposed on the original image. However, unsharp masking has the deficiency that the noise is also enhanced due to its principle of operation when noise is superposed on the blurred image. In order to alleviate this problem, this paper proposes a characteristic quantity of image edges (edge information) using the amplitude shift between the scales of the multiscale brightness gradient, and applies it to the control of the emphasis of unsharp masking. The proposed edge information takes a positive value on the image edge and the amplitude is proportional to the smoothness and brightness difference of the edges. Since the proposed edge information takes a positive value at the smoothed edges regardless of the variance of the superposed Gaussian noise and the image contrast, it is possible to reduce the effect of noise in the emphasis result by controlling the image emphasis by discriminating positive or negative values. In the experiment, the proposed edge information is introduced into unsharp masking using a fuzzy rule and into third-order unsharp masking. The effectiveness of the proposed edge information is confirmed by image emphasis.

A cascade configuration of the edge‐weighted image enhancement filter

“…9, a nonseparable high-pass filter is used. However, the design of the weight functions by analysis of the noise amplification characteristics presented in this paper can be applied to the edge weight enhancement filters [4,8] based on a nonseparable high-pass filter. It is a future topic of study to introduce the proposed weighted high-pass filter for use in image resolution enhancement.…”

“…This problem is caused by the fact that the variations of the edge weight function and the high-pass filter output do not coincide near the edge to be emphasized. Also, in the edge-weighted enhancement filter using a nonseparable high-pass filter [4,7,8], the same problem as that in the SCUM method occurs because the amplitude of the high-pass filter approaches 0 at the center of the brightness variations. In the next section, the second-order differential used in the SCUM method is decomposed into two first-order differentials in the proposed configuration of a new edge-weighted high-pass filter in which noise amplification on the edges is suppressed.…”

“…The difference used is that defined at the same center coordinates as the one for the impulse response of the high-pass filter in the previous stage [Eqs. (8) and (9)], so that the maximum of the absolute value of the amplitude of the output from the high-pass filter in the previous stage appears at the same coordinates with respect to the edge. Then, the weight functions are either or in which the distance between the pixels used in computing the differential is three pixels.…”

“…Here α 1 and α 2 are constants. Like the present weight functions, there is a method that uses the edge characteristic values obtained from several scales of the discrete binary wavelet transform of the signal [4,8]. 21) and (22) and then setting coefficient values to minimize the square error between the high-pass filter output and the weighted high-pass filter output.…”

“…In Ref. 8, the weight function is defined in terms of the differential of the wavelet transform. In each method, an optimum weight function in the sense of minimizing the noise amplification is not derived.…”