Noise reduction apparatus and noise reduction method

1. A noise reduction apparatus comprising:

an input section which inputs an image signal;

a plurality of first high pass filters which receive the image signal input to the input section and which have different passing frequency bands;

a plurality of level histogram generation sections which are provided so as to correspond to the plurality of first high pass filters and each of which generates a level histogram representing the number of pixels of the image signal counted for each amplitude level thereof;

a detection section which detects one of the level histograms generated by the plurality of level histogram generation sections that satisfy a predetermined threshold value;

a first selection section which selects the output of the first high pass filter that has supplied its output to the level histogram generation section that has generated the level histogram detected by the detection section;

a first noise reduction section which applies high frequency component noise reduction processing to the output of the first high pass filter selected by the first selection section;

a first subtractor which subtracts the output of the first high pass filter selected by the first selection section from the image signal input to the input section;

a second noise reduction section which applies low frequency component noise reduction processing to the output of the first subtracter; and

an adder which adds outputs of the first and second noise reduction sections.

2. The noise reduction apparatus according to claim 1 , further comprising:

a delay section which delays a noise reduction processed image signal output from the adder on a per frame basis;

a plurality of second high pass filters which receive the frame-delayed image signal from the delay section and which have the same passing frequency bands as those of the plurality of first high pass filters, respectively;

a second selection section which selects, based on a detection result of the detection section, an output of the second high pass filter having the same passing frequency band as that of the first high pass filter selected by the first selection section; and

a second subtractor which subtracts an output of the second high pass filter selected by the second selection section from the frame-delayed image signal received from the delay section, wherein

the first noise reduction section is a motion adaptive noise reduction circuit that receives the output of the first high pass filter selected by the first selection section and output of the second high pass filter selected by the second selection section, and

the second noise reduction section is a motion adaptive noise reduction section that receives the outputs of the first and second subtractors.

3. The noise reduction apparatus according to claim 1 , further comprising:

a delay section which delays an image signal input to the input section on a per frame basis;

a plurality of second high pass filters which receive the frame-delayed image signal from the delay section and which have the same passing frequency bands as those of the plurality of first high pass filters, respectively;

a second selection section which selects, based on a detection result of the detection section, an output of the second high pass filter having the same passing frequency band as that of the first high pass filter selected by the first selection section; and

a second subtractor which subtracts an output of the second high pass filter selected by the second selection section from the frame-delayed image signal received from the delay section, wherein

the first noise reduction section is a motion adaptive noise reduction circuit that receives the output of the first high pass filter selected by the first selection section and output of the second high pass filter selected by the second selection section, and

the second noise reduction section is a motion adaptive noise reduction section that receives the outputs of the first and second subtractors.

4. The noise reduction apparatus according to claim 1 , wherein

the detection section detects one of the level histograms generated by the respective histogram generation sections, in which the number of pixels having an amplitude level exceeding a first threshold value corresponds to the number of those exceeding a second threshold value.

5. The noise reduction apparatus according to claim 1 , wherein

the detection section determines whether the level histogram that the level histogram generation section has generated based on the output of the first high pass filter having the highest passing frequency band of the plurality of first high pass filters satisfies a predetermined threshold value and, if not satisfy, determines whether the level histogram that the level histogram generation section has generated based on the output of the first high pass filter having a passing frequency band one step lower than the first high pass filter having the highest passing frequency band satisfies a predetermined threshold value.

6. The noise reduction apparatus according to claim 1 , wherein

the first high pass filter configured to be a horizontal high pass filter, a vertical high pass filter, or horizontal/vertical high pass filter.

7. An image output apparatus comprising:

an input section which inputs an image signal;

a plurality of first high pass filters which receive the image signal input to the input section and which have different passing frequency bands;

a plurality of level histogram generation sections which are provided so as to correspond to the plurality of first high pass filters and each of which generates a level histogram representing the number of pixels of the image signal counted for each amplitude level thereof;

a detection section which detects one of the level histograms generated by the plurality of level histogram generation sections that satisfies a predetermined threshold value;

a first selection section which selects an output of the first high pass filter that has supplied its output to the level histogram generation section that has generated the level histogram detected by the detection section;

a first noise reduction section which applies high frequency component noise reduction processing to the output of the first high pass filter selected by the first selection section;

a first subtractor which subtracts the output of the first high pass filter selected by the first selection section from the image signal input to the input section;

a second noise reduction section which applies low frequency component noise reduction processing to an output of the first subtractor;

an adder which adds outputs of the first and second noise reduction sections; and

an output section which outputs the noise reduction processed image signal received from the adder.

8. A noise reduction method comprising:

inputting an image signal;

supplying the input image signal to a plurality of first high pass filters which have different passing frequency bands;

supplying outputs of the plurality of first high pass filters, in a corresponding manner, to a plurality of level histogram generation sections each of which generates a level histogram representing the number of pixels of the image signal counted for each amplitude level thereof;

detecting one of the level histograms generated by the plurality of level histogram generation sections that satisfies a predetermined threshold value;

selecting an output of the first high pass filter that has supplied its output to the level histogram generation section that has generated the detected level histogram;

applying high frequency component noise reduction processing to the output of the selected first high pass filter;

subtracting the output of the selected first high pass filter from the image signal input to the input section;

applying low frequency component noise reduction processing to the output obtained by the subtraction; and

adding outputs obtained by the high frequency component noise reduction processing and low frequency component noise reduction processing.

9. The noise reduction method according to claim 8 , wherein

the detecting detects whether the level histogram that the level histogram generation section has generated based on the output of the first high pass filter having the highest passing frequency band of the plurality of first high pass filters satisfies a condition that the number of pixels having an amplitude level exceeding a first threshold value corresponds to the number of those exceeding a second threshold value and, if not satisfy, detecting whether the level histogram that the level histogram generation section has generated based on the output of the first high pass filter having a passing frequency band one step lower than the first high pass filter having the highest passing frequency band satisfies a condition that the number of pixels having an amplitude level exceeding a first threshold value corresponds to the number of those exceeding a second threshold value.