Noise reduction method and device for image signal

Abstract

The first difference signal between the current and previous frame signals is obtained by the subtracter pixel by pixel and supplied to both the line memory and the average value calculation circuit. The average value calculation circuit calculates the average value for at least one line of the first difference signal which relates to only pixels with smaller absolute of the difference value than the present threshold. If the pixels which have absolute values smaller than the threshold, are less than the present number, the average value is set 0. Large values of the first difference caused by motion of the target are excluded and not treated as the flicker noise. The second difference signal is transformed by the non-linear circuit according to predetermined characteristics and supplied to the adder. The previous frame signal is supplied to the adder at the same time by delaying the signal through the line memory. The adder adds the output signals of the non-linear circuit and the line memory at each same pixel, and outputs to the output port and the frame memory. In a second embodiment the noise reduction device may include a noise reduction circuit and evaluation calculation circuit to decrease signal distortion for more natural image quality. In a third embodiment the noise reduction device may include a discrimination circuit and selectively switchable multipliers to adjust factors to optimize noise reduction or distortion reduction based on human visual sensitivity.