JP2008099196A - Signal processing method, signal processing circuit and imaging apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a signal processing method, signal processing circuit and imaging apparatus in which color noise is reduced without damaging a color tone of an overall image. <P>SOLUTION: A low saturation compressing unit 532 sets non-compressibility (r) in accordance with a luminance signal Y and composes, in accordance with the non-compressibility (r), a compressed color difference signal obtained by compressing a color difference signal of a fixed level th-cmp or lower and a non-compressed color difference signal. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a signal processing method, a signal processing circuit, and an imaging apparatus that suppress color noise.

In an imaging device such as a digital camera using an imaging device that converts incident light into an electrical signal and generates image data, noise is particularly generated in the generated image data, particularly when the imaging sensitivity is increased. .
Noise can be broadly classified into dot-like luminance noise generated in the luminance signal and patchy color noise generated in the color signal. Among these, color noise is random noise of a low frequency, and causes significant image quality degradation in the generated image data.

  If such a color noise is to be removed with a linear filter, a very multi-tap filter is required, which is expensive and undesirable. For this reason, a method for removing color noise without much cost has been considered.

For example, there is a technique called low saturation compression.
This is a method of reducing color noise by compressing a low-saturation portion in image data and eliminating color.
In addition, there is a method disclosed in Patent Documents 1 and 2 in which the saturation is reduced according to the luminance of each part in the image data.
JP 2001-189945 A JP 2003-235055 A

However, the above-described technique of low saturation compression has the disadvantage that the nuance of the entire image may change when applied to the entire image data, and is actively applied to reduce color noise. I couldn't.
Further, in the methods disclosed in Patent Documents 1 and 2 described above, the saturation of the entire image data is lowered according to the luminance value of each part of the image data. There was a disadvantage that became faint.

  In order to eliminate the disadvantages described above, an object of the present invention is to provide a signal processing method, a signal processing circuit, and an imaging apparatus that reduce color noise without impairing the color of the entire image.

  In order to achieve the above-described object, a signal processing method according to a first aspect of the present invention includes a first step of setting a predetermined ratio of 0 or more and 1 or less according to a luminance value of a luminance signal separated from image data. The color signal relating to the color of the image data is compressed in a region where the absolute value of the input level of the color signal is not more than a first threshold value, and the absolute value of the input level of the color signal is set to the first threshold value. A second step of generating a compressed color signal that is not compressed with respect to a larger region, the uncompressed color signal, and the compressed color signal generated in the second step are And a third step of generating a synthesized color signal synthesized in accordance with the predetermined ratio set in the first step.

  The signal processing circuit of the second invention sets a predetermined ratio of 0 or more and 1 or less according to the luminance value of the luminance signal separated from the image data, and inputs a color signal related to the color of the image data to the color signal. A compressed color signal is generated for a region where the absolute value of the level is equal to or less than the first threshold value, and is not compressed for a region where the absolute value of the input level of the color signal is greater than the first threshold value. Then, a combined color signal is generated by combining the uncompressed color signal and the generated compressed color signal in accordance with the set predetermined ratio.

  An imaging apparatus according to a third aspect of the present invention includes an imaging unit that generates image data, and a signal processing unit that performs signal processing on the image data generated by the imaging unit. A signal separation unit that separates the image data generated by the imaging unit into a luminance signal and a color signal relating to color, and a predetermined ratio of 0 to 1 in accordance with the luminance value of the luminance signal separated by the signal separation unit A region where the absolute value of the input level of the color signal is compressed with respect to a region where the absolute value of the input level of the color signal is equal to or less than the first threshold, and the absolute value of the input level of the color signal is greater than the first threshold A compressed color signal that is not compressed is generated, and a combined color signal obtained by combining the uncompressed color signal and the generated compressed color signal in accordance with the set predetermined ratio. A color signal processing unit to generate

  According to the present invention, it is possible to provide a signal processing method, a signal processing circuit, and an imaging apparatus that reduce color noise without impairing the color of the entire image.

The configuration of the camera system 100 (corresponding to the imaging apparatus of the present invention) to which the motion vector detection method of the present invention is applied will be described below.
The camera system 100 of FIG. 1 includes, for example, a lens 1, an image sensor 2, a preprocessing unit 3 having a CDS (correlated double sampling), a preamplifier, and the like, an AD (analog / digital) converter 4, and a distortion defect processing unit 51. , A luminance signal processing unit 52, a signal processing unit 5 having a color signal processing unit 53, and the like.

  The lens 1 projects an image of a subject (not shown) on the imaging surface of the image sensor 2. The imaging device 2 is composed of, for example, a CCD (Charge Coupled Device), a CMOS (Complementary Metal Oxide Semiconductor), and the like, converts an image transmitted through the lens 1 into an electrical signal, and uses CDS (correlated double sampling) or a preamplifier as an image signal. To the pre-processing unit 3 having

  The preprocessing unit 3 samples and holds the pixel signal from the image pickup device 2 to extract necessary data, performs gain control (AGC: Automatic Gain Control) to adjust to an appropriate level, and also performs black level adjustment. Is called. The output signal of the preprocessing unit 3 is output to the AD converter 4 at the subsequent stage.

  Since the AD converter 4 handles a pixel (image) signal, one having an accuracy of 10 to 12 bits is adopted. The AD converter 4 converts the output signal supplied from the preprocessing unit 3 from an analog signal to a digital signal, and outputs the digital signal to the signal processing unit 5 connected to the subsequent stage.

The signal processing unit 5 includes a distortion defect processing unit 51, a luminance signal processing unit 52, a color signal processing unit 53, and the like, and performs digital signal processing.
In the distortion defect processing unit 51, black detection, digital gain adjustment (Digital Gain Control), shading correction generated in the lens 1, luminance signal and color signal are separated using a delay line, and pixel defect is detected using a delay line. Perform the correction.

The luminance signal processing unit 52 performs Y (luminance) signal vertical / horizontal (direction) contour correction, Y luminance signal and vertical / horizontal contour correction signal Mix (mixing; addition) processing, γ (gamma) correction, luminance (luminance). ) Various image processing such as key processing, solarization processing for processing a partial area of the image to an arbitrary luminance, and negative processing for inverting the image.
The color signal processing unit 53 performs color separation and clamping processing, removal of color signal noise and color false signals, RGB matrix (Matrix) processing, white balance (WB) adjustment for varying the coefficients of R, G, and B colors, γ (gamma) correction, RG / BG conversion, color false signal suppression processing, color difference signal (Cr / Cb) generation, chroma suppression (suppression) processing, Hue / Gain adjustment, and the like are performed.

The color signal processing unit 53 further includes a color difference signal generation unit 531 and a low saturation compression unit 532.
The color difference signal generation unit 531 generates color difference signals Cr and Cb based on the input red signal R, green signal G, and blue signal B.
The low saturation compression unit 532 performs low saturation compression processing described later based on the color difference signal Cr / Cb generated by the color difference signal generation unit 531 and the luminance signal Y input from the luminance signal processing unit.

Next, the low saturation compression process performed by the low saturation compression unit 532 will be described.
In the conventional low-saturation compression process, as described in the background art and problem, a portion with low saturation in the image data is compressed, and color noise is reduced by eliminating the color. For this reason, a light hue close to the hue of the color noise is compressed, resulting in a disadvantage that the nuances of the entire image data are different.
More specifically, when the input color difference signal Cr or Cb is below a certain level (for example, threshold value th_cmp), the output value is compressed as shown in FIG. That is, the portion with high saturation (the large color difference signal) is output as it is, but the portion with low saturation (the color difference signal of th_cmp or less) is compressed and approaches an achromatic color, so that the color noise is reduced. At the same time, the nuances of the entire image are different.
The threshold th_cmp is, for example, a value set empirically, and the setting method is not limited in the present invention.

  Therefore, in the low saturation compression unit 532 of the present embodiment, after performing the conventional low saturation compression process described above, the original uncompressed color difference signal and the compressed color difference signal (hereinafter referred to as the compressed color difference signal Cr_cmp). And Cb_cmp) at a suitable ratio, it is possible to suppress the compression of the portion related to the color of the entire image while compressing the color noise portion.

  When the uncompressed color difference signals Cr and Cb and the compressed color difference signals Cr_cmp and Cb_cmp are combined, they are combined according to the following equation (1) to generate combined color difference signals Cr_com and Cb_com.

The uncompressed ratio r (predetermined ratio) is a coefficient indicating a ratio when the uncompressed color difference signal and the compressed color difference signal are combined, and is calculated based on the luminance signal Y. A method for calculating the non-compression rate r will be described later.
In practice, if the equation (1) is modified into the equation shown in the following equation (2), the circuit scale can be suppressed as compared with the equation (1).

  Although the color difference signal Cr is shown in the above formulas (1) and (2), the color difference signal Cb is also synthesized by the same formula.

In the above formulas (1) and (2), the coefficient r takes a value from 0 to 1.
When the uncompressed ratio r is 0, the output is Cr_cmp, which means that the output signal is only a completely compressed signal.
Conversely, when the uncompressed ratio r is 1, the output is Cr, which means that the output signal is only a signal that is not compressed at all.

Next, a method for setting the value of the non-compression rate r will be described.
FIG. 3 is a diagram for explaining a method for setting the non-compression rate r.
In FIG. 3, the vertical axis represents the uncompressed ratio r, and the horizontal axis represents the luminance value Yin of the luminance signal.
As shown in FIG. 3, when the luminance value Yin is less than a certain threshold th_start, the non-compression rate r is set to zero.
When the luminance value Yin is greater than or equal to the threshold th_start and r is less than 1, the uncompressed rate r is given by the following equation (3).

Here, gain is a value representing the slope of the non-compression rate r, and can be set freely. If the slope gain is set to a small value, the uncompressed ratio r changes gently according to the luminance value Yin.
When the uncompressed rate r reaches 1 according to the equation (3), the uncompressed rate r is maintained at 1 for the luminance value Yin higher than that.

  The threshold value th_start and the slope gain described above need to be appropriately adjusted so that the output color difference signals Cr and Cb have optimum color difference values. That is, the values of the threshold th_start and the slope gain vary depending on the input image data. For example, the user changes the threshold th_start and the slope gain while comparing the input image data and the output image data. The threshold value th_start and the inclination gain are adjusted so that a desired output image is output.

Hereinafter, a specific example of the adjustment of the threshold th_start and the inclination gain will be described.
FIG. 4 is a diagram for explaining an example of setting of the threshold th_start and the slope gain.
4A is a diagram showing the relationship between the input signal amount to the image sensor 2 and the gain value applied to the input signal, and FIGS. 4B to 4D show the gain values in FIG. It is a figure which shows the specific example which sets threshold value th_start and inclination gain according to it.
Since the S / N ratio decreases (noise increases) as the gain value applied to the image sensor 2 increases, the luminance range for executing the saturation compression is set according to the gain value shown in FIG. As a result, it is possible to obtain an image that is appropriately chroma-compressed.

That is, when the gain value shown in FIG. 4A is small (gain values 1 to 9), it is determined that the S / N ratio is not so bad, and as shown in FIG. The compression ratio r is fixed to 1 (that is, no saturation compression is performed), and the input signal is output as it is.
For example, when the gain value is large (when the gain value is 18 to 36), it is determined that the S / N ratio is bad, and as shown in FIG. 4D, for example, th_start is 5%, and the slope gain is 3 By setting as above, the saturation compression range can be increased.
Further, in the case where the gain value is intermediate between the case where the gain value is large and the case where the gain value is small (when the gain value is 9 to 18), as shown in FIG. 4C, the non-compression ratio r is also intermediate between the above two cases. For example, the threshold th_start is set to 0% and the slope gain is set to 8.

By setting in this way, it is possible to determine that the S / N ratio is large, increase the ratio of saturation compression at a portion where the gain value is large, and increase the saturation ratio at a portion where the S / N ratio is small and determine that the S / N ratio is small. By reducing the compression ratio, saturation compression can be appropriately performed on the input image, and color noise can be reduced without impairing the color of the entire image.
Note that the above-described method for setting the threshold th_start and the slope gain is an example, and the non-compression rate r of the present invention is not limited to the above-described setting method. For example, it may be determined in advance at the time of designing the camera system 100 and stored at the time of shipment, or the user may set in real time while comparing the input image and the output image.

Next, an operation example of the camera system 100 of the present embodiment will be described.
FIG. 5 is a flowchart for explaining an operation example of the camera system 100.
Step ST1:
Based on the light incident on the lens 1, the image sensor 2 generates image data.
Step ST2:
The image data generated in step ST1 is input to the signal processing unit 5, and the distortion defect processing unit 51 separates the luminance signal Y and the color signals R, G, and B.

Step ST3:
The color difference signal generation unit 531 generates color difference signals Cr and Cb based on the color signals R, G, and B.
Step ST4:
The non-compression rate r is set by a user operation or the like via an operation unit (not shown in FIG. 1).
The non-compression ratio r may be set by a user operation, or may be set and stored in advance according to a gain value or the like. In the operation example shown in FIG. 5, the non-compression rate r is set in this step ST4. However, the present invention is not limited to this, and the setting of the non-compression rate r is low in step ST6 described later. It may be performed before the processing performed by the degree compression unit 532.

Step ST5:
The low saturation compression unit 532 generates compressed color difference signals Cr_cmp and Cb_cmp by compressing Cr and Cb whose color difference signals are equal to or less than th_cmp based on the color difference signals Cr and Cb generated in step ST3.
Step ST6:
The low saturation compression unit 532 uses the compressed color difference signals Cr_cmp and Cb_cmp generated in step ST5 and the uncompressed color difference signals Cr and Cb using the value of the uncompressed ratio r set in step ST4. The synthesized color difference signals Cr_com and Cb_com are generated by synthesizing according to the formula (1) or (2).
Step ST7:
The signal processing unit 5 converts the combined color difference signals Cr_com and Cb_com and the luminance signal Y into analog signals and outputs them.

  As described above, according to the camera system 100 of the present embodiment, the uncompressed rate r is set according to the luminance signal Y, and the compressed color difference signal obtained by compressing the color difference signal of a certain level th_cmp or less is compressed. By combining the non-compressed color difference signal according to the uncompressed ratio r, in the low luminance part where the color noise is conspicuous, the ratio of the compressed color difference signal is increased and the color noise is strongly removed, and the uncompressed ratio is appropriately set. By setting to, it is possible to avoid the fact that the color of the entire image is different.

In addition, in the conventional circuit that can execute the low saturation compression process, a multi-tap filter is necessary. However, according to the camera system 100 of the present embodiment, the circuit configuration using Expression (2) is adopted. It becomes easy to configure the circuit.
Further, since the non-compression rate r can be set freely, appropriate color noise removal can be performed according to input image data.

The present invention is not limited to the embodiment described above.
That is, when implementing the present invention, various modifications, combinations, sub-combinations, and alternatives may be made to the components of the above-described embodiments within the technical scope of the present invention or an equivalent scope thereof.

  In the embodiment described above, the image data is separated into the luminance signal and the color signal (RGB) signal, the color difference signals Cr and Cb are generated based on the RGB signal, and the low saturation compression and synthesis are performed using the color difference signal. Although the camera system 100 that performs processing has been described, the present invention is not limited to this. That is, in the present invention, for example, an R-G / B-G / G signal or the like may be generated from an RGB signal, and the above-described processing may be performed based on this. Further, the above-described processing may be performed by separating the color signal into another color space.

  In the above-described embodiment, the example in which the two signals of the compressed color difference signal and the uncompressed color difference signal are combined has been described. However, the present invention is not limited to this. That is, in the present invention, the region to be compressed in the signal to be compressed (the portion below the input level th_cmp) and its compression rate are divided into two or more according to the input level, and two or more compressed by different compression rates, respectively. These color difference signals and the uncompressed color difference signals may be combined at a ratio corresponding to a preset non-compression rate r ′.

FIG. 1 is a block diagram showing the configuration of the camera system 100. FIG. 2 is a diagram for explaining low saturation compression. FIG. 3 is a diagram for explaining a method for setting the non-compression rate r. FIG. 4 is a diagram for explaining an example of setting of the threshold th_start and the slope gain. FIG. 5 is a flowchart for explaining an operation example of the camera system 100.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 ... Camera system, 1 ... Lens, 2 ... Image sensor, 3 ... Pre-processing part, 4 ... AD converter, 5 ... Signal processing part, 51 ... Distortion defect processing part, 52 ... Luminance signal processing part, 53 ... Color signal Processing unit, 531 ... Color difference signal generation unit, 532 ... Low saturation compression unit

Claims (4)

A first step of setting a predetermined ratio of 0 or more and 1 or less according to the luminance value of the luminance signal separated from the image data;
The color signal relating to the color of the image data is compressed in a region where the absolute value of the input level of the color signal is equal to or less than the first threshold value, and the absolute value of the input level of the color signal is less than the first threshold value. A second step of generating a compressed color signal that is not compressed for a large region;
A combined color signal is generated by combining the uncompressed color signal and the compressed color signal generated in the second step according to the predetermined ratio set in the first step. A third step;
A signal processing method. In the third step, the uncompressed color signal is multiplied by the predetermined ratio set in the first step, and the compressed color signal generated in the second step is (1- (The predetermined ratio) is multiplied by the coefficient, and the uncompressed color signal multiplied by the predetermined ratio is added to the compressed color signal multiplied by the coefficient of (1-the predetermined ratio). The signal processing method according to claim 1, wherein a color signal is generated. A predetermined ratio of 0 or more and 1 or less is set according to the luminance value of the luminance signal separated from the image data, and the absolute value of the input level of the color signal of the color signal related to the color of the image data is the first threshold. A compressed color signal that is compressed with respect to a region that is less than or equal to a value, and that is not compressed with respect to a region in which the absolute value of the input level of the color signal is greater than the first threshold, A signal processing circuit that generates a composite color signal by combining the generated color signal to be compressed in accordance with the set predetermined ratio. An imaging unit for generating image data;
A signal processing unit that performs signal processing on the image data generated by the imaging unit, and
The signal processing unit
A signal separation unit for separating the image data generated by the imaging unit into a luminance signal and a color signal relating to color;
A predetermined ratio of 0 or more and 1 or less is set according to the luminance value of the luminance signal separated by the signal separation unit, and the absolute value of the input level of the color signal is equal to or smaller than a first threshold value. A compressed color signal is generated that is compressed with respect to a region, and is not compressed with respect to a region where the absolute value of the input level of the color signal is greater than the first threshold, and the color signal that is not compressed is generated. A color signal processing unit that generates a composite color signal obtained by combining the compressed color signal with the set predetermined ratio;
An imaging apparatus having

Images