JP6272115B2 - Endoscope processor and endoscope system - Google Patents

Description

  The present invention relates to an endoscope image processing apparatus and an endoscope system that process pixel signals of respective pixels of a subject imaged by an imaging element.

  An endoscope system for observing the inside of a lumen such as a human esophagus or intestine is known. This type of endoscope system includes an endoscope processor that processes a pixel signal of each pixel of a subject imaged by an electronic scope. The endoscope processor performs image processing such as color conversion processing and noise reduction processing on the pixel signal in order to display an observation image that is easy to see for the operator on the monitor.

  For example, Patent Document 1 describes an image processing apparatus that performs noise reduction processing using component separation on a pixel signal. In the image processing apparatus described in Patent Document 1, an image signal is separated into two components, a skeleton component including edge information and a residual component including a noise component, and noise reduction processing parameters are obtained using the separated skeleton components. Is set. Further, a correlation coefficient between color components of the pixel signal is calculated for the pixel of interest and its surrounding pixels, and the noise reduction processing parameter is corrected based on the correlation coefficient. Using this corrected noise reduction processing parameter, noise reduction processing is performed on the residual component. By synthesizing the skeleton component and the residual component that has been subjected to noise reduction processing, an image signal with reduced noise is obtained.

JP 2010-166598 A

  In the image processing apparatus described in Patent Document 1, in order to perform noise reduction processing on an image signal, the image signal is separated into a skeleton component and a residual component. The separation of the image signal components is complicated in calculation formula and requires a large calculation load. Further, in order to correct the noise reduction processing parameter, a correlation coefficient between the color components of the pixel signal is calculated for the pixel of interest and its surrounding pixels. The calculation amount of the correlation coefficient increases remarkably as the number of surrounding pixels increases. In order to implement noise reduction processing using component separation and correlation coefficients with complicated calculation formulas on hardware, a lot of hardware resources are required, making it difficult to commercialize.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to perform endoscopic image processing suitable for reducing noise (improvement of the SN ratio) of pixel signals with a small amount of hardware resources. An apparatus and an endoscope system are provided.

  In order to achieve the above object, an endoscope image processing apparatus according to an embodiment of the present invention is an apparatus that processes pixel signals of each pixel of a subject imaged by a predetermined imaging element. Component separation means for separating the pixel signal of the target pixel into a luminance component and a color component, and parameter determination means for determining a predetermined parameter relating to the target pixel based on at least one of the separated luminance component and color component. Noise reduction means for reducing the noise of the pixel of interest by applying a predetermined spatial filtering process to the pixel of interest based on the parameters.

  According to such a configuration, the parameter used for the spatial filter processing is determined based on the pixel signal of the target pixel. Therefore, it is not necessary to perform complicated calculation using peripheral pixels to determine the parameters, and the noise of the pixel signal can be reduced (the SN ratio is improved) with less hardware resources.

  Further, the noise reduction means obtains a pixel signal of the target pixel by taking a weighted average of the pixel signals for each color component for the target pixel and peripheral pixels around the target pixel based on the parameter determined by the parameter determination means. Spatial filter processing may be applied to.

  Further, the parameters may include a size parameter indicating the number of peripheral pixels from which a weighted average of pixel signals is taken and a position of the peripheral pixel with respect to the target pixel, and an intensity parameter indicating a weight of each pixel signal from which the weighted average is taken. .

  The parameter determination unit may calculate an average value of all the pixel signals for each color component and determine the parameter based on the calculated average value.

  The parameter determining means may calculate the difference between the luminance component and the color component for each pixel, and determine the parameter based on the calculated difference.

  The parameter determining means may calculate an average value of all the pixel signals for each color component, and determine one of the size parameter and the intensity parameter based only on the calculated average value.

  According to such a configuration, one of the parameters used for the spatial filter processing is determined for each color component based on only the average value of the pixel signals of all pixels. Therefore, it is not necessary to perform complicated calculation using peripheral pixels to determine the parameters, and the noise of the pixel signal can be reduced (the SN ratio is improved) with less hardware resources.

  The parameter determination unit may calculate the difference between the luminance component and the color component for each pixel and determine the other of the size parameter and the intensity parameter based only on the calculated difference.

  According to such a configuration, for each pixel, one of the parameters used for the spatial filter processing is determined based only on the pixel signal of the target pixel. Therefore, it is not necessary to perform complicated calculation using peripheral pixels to determine the parameters, and the noise of the pixel signal can be reduced (the SN ratio is improved) with less hardware resources.

  An endoscope system according to an embodiment of the present invention includes the above-described endoscope image processing apparatus and an electronic scope having a predetermined imaging device.

  According to the embodiment of the present invention, an endoscopic image processing apparatus and an endoscopic system suitable for reducing noise of a pixel signal (improvement of the SN ratio) with a small amount of hardware resources are provided.

1 is a block diagram of an endoscope system according to an embodiment of the present invention. It is the figure which represented typically a part of light-receiving surface of the image pick-up element concerning embodiment of this invention. It is the figure which represented typically a part of light-receiving surface of the image pick-up element concerning embodiment of this invention. It is a flowchart which shows the operation | movement flow of the endoscope system concerning embodiment of this invention. It is the figure which represented typically a part of light-receiving surface of the image pick-up element concerning embodiment of this invention.

  Hereinafter, an endoscope system according to an embodiment of the present invention will be described.

  FIG. 1 is a block diagram showing a configuration of an endoscope system 1 of the present embodiment. As shown in FIG. 1, the endoscope system 1 is a medical imaging system, and includes an electronic scope 100, an endoscope processor 200, and a monitor 300.

  The electronic scope 100 includes an objective optical system 101, an image sensor 102, an image sensor driver 103, an AFE (Analog Front End) 104, an illumination optical system 105, and a light guide 106. The objective optical system 101, the image sensor 102, and the illumination optical system 105 are provided in the distal end portion 100 a of the electronic scope 100.

  The endoscope processor 200 incorporates an endoscope image processing apparatus, and includes a system controller 201, a timing controller 202, a light source 203, a light source driver 204, an image processing unit 205, an image memory 206, and a video signal generation circuit 207. And a front panel 208. The image processing unit 205 includes a demosaic processing circuit 205A, a color adjustment circuit 205B, and a noise removal circuit 205C. The endoscope processor 200 is preferably configured to photograph a subject in which a specific color component is dominant (for example, a body cavity of a person in which red is dominant).

  The system controller 201 controls each element constituting the endoscope system 1. The timing controller 202 transmits a clock pulse for adjusting a signal processing timing to each circuit in the endoscope system 1.

  The light source 203 is driven and controlled by the light source driver 204 to emit white light. As the light source 203, a high-intensity lamp such as a xenon lamp, a halogen lamp, a mercury lamp, or a metal halide lamp is used. The illumination light emitted from the light source 203 is incident on the light guide 106 and guided in the light guide 106 toward the distal end portion 100 a of the electronic scope 100.

  The illumination light guided in the light guide 106 is emitted from the end face of the light guide 106 disposed in the distal end portion 100a. Illumination light emitted from the end face of the light guide 106 is emitted from the distal end portion 100a via the illumination optical system 105 and illuminates the subject. Illumination light (reflected light) reflected by the subject enters the image sensor 102 via the objective optical system 101 and forms a subject image on the light receiving surface of each pixel included in the image sensor 102.

  The image sensor 102 includes R, G, and B pixels having red (R), green (G), and blue (B) color filters, respectively. Each pixel accumulates the formed subject image as a charge corresponding to the amount of light, and converts it into pixel signals (R pixel signal, G pixel signal, B pixel signal) corresponding to each color of R, G, B. Each converted pixel signal is subjected to signal amplification processing and A / D conversion processing by the AFE 104 and transmitted to the image processing unit 205 of the endoscope processor 200. For example, a CCD (Charge Coupled Device) image sensor or a CMOS (Complementary Metal Oxide Semiconductor) image sensor is used as the image sensor 102.

  The R, G, and B pixel signals received by the image processing unit 205 are stored in the image memory 206. The image processing unit 205 performs image processing on each pixel signal on the work space in the image memory 206. The image processing includes demosaic processing by the demosaic processing circuit 205A, color adjustment processing by the color adjustment circuit 205B, and noise removal processing by the noise removal circuit 205C.

  The demosaic processing circuit 205A performs demosaic processing on each pixel signal. The pixel signal of each pixel input to the demosaic processing circuit 205A has only one color information among R, G, and B. In the demosaic process, a process of giving information on all the colors R, G, and B to the pixel signal of each pixel is performed.

  The demosaic process will be described with reference to FIG. FIG. 2 schematically shows a part of the light receiving surface of the image sensor 102. In FIG. 2, “R”, “G”, and “B” indicate R, G, and B pixels, respectively. In the demosaic process for the pixel G1, an interpolation process using pixel signals output from R pixels (for example, the pixel R1 and the pixel R2) arranged around the pixel G1 is performed. In addition, interpolation processing is performed using pixel signals output from B pixels (for example, pixel B1 and pixel B2) arranged around the pixel G1. By adding the R pixel signal and the B pixel signal obtained by this interpolation processing to the pixel signal of the pixel G1, the pixel signal of the pixel G can have information on all the colors of R, G, and B. Similarly, by performing interpolation processing using the pixel signals of surrounding pixels for all pixels, all the pixels can be provided with information on all colors of R, G, and B. The pixel signal that has been demosaiced is transmitted to the color adjustment circuit 205B.

  The color adjustment circuit 205B performs color adjustment processing on the pixel signal transmitted from the demosaic processing circuit 205A. The color adjustment processing includes, for example, known matrix calculation processing, white balance adjustment processing, gamma correction processing, and the like. The pixel signal subjected to the color adjustment processing is transmitted to the noise removal circuit 205C. The noise removal circuit 205C performs a later-described noise removal process (SN ratio improvement process) on the pixel signal received from the color adjustment circuit 205B. The pixel signal that has been subjected to noise removal processing is transmitted to the video signal generation circuit 207.

  The video signal generation circuit 207 converts the pixel signal received from the image processing unit 205 (noise removal circuit 205C) into a video signal of a predetermined format (for example, NTSC format), and transmits the video signal to the monitor 300. The monitor 300 displays an observation image based on the video signal received from the video signal generation circuit 207.

  Next, noise removal processing by the noise removal circuit 205C will be described in detail. In the noise removal process, a spatial filter process is performed on the pixel signal of each pixel. FIG. 3 schematically shows a part of the light receiving surface of the image sensor 102 and is a diagram for explaining an example of the spatial filter processing. In FIG. 3, the case where the spatial filter process is performed on the pixel signal of the target pixel P1 surrounded by the thick solid line will be described. In the spatial filter process for the pixel signal of the target pixel P1, a weighted average value of the pixel signals of the target pixel P1 and its surrounding pixels is calculated. In the example shown in FIG. 3, a weighted average value of pixel signals of 3 × 3 nine surrounding pixels centered on the target pixel P1 is calculated. The numerical value described in each pixel in FIG. 3 represents the weight in the weighted average value. The weight of the pixel signal of the target pixel P1 is the largest, and then the weight of the pixel signal of the peripheral pixels that are in contact with the target pixel P1 either vertically or horizontally is large. Moreover, the weight of the pixel signal of the peripheral pixel arrange | positioned at four corners among nine pixels is the smallest. In addition, the pixel signal of the pixel in which the numerical value is not described is not used for calculation of the weighted average value. Hereinafter, for convenience of explanation, the range of pixels for which the weighted average value is calculated and the weight of the pixel signal of each pixel are collectively referred to as a “spatial filter”. When the spatial filter process is performed on the pixel signal of the target pixel P1, a weighted average value is calculated based on the range and weight of the pixel defined by the spatial filter, and the pixel signal of the target pixel P1 is changed to the weighted average value. Replaced.

  The pixel signal of each pixel includes three color components (R component, G component, and B component) by demosaic processing by the demosaic processing circuit 205A. For this reason, the spatial filter processing is executed independently for each color component. Further, the spatial filter processing is executed independently for the pixel signals of all the pixels.

  The number and arrangement of pixels defined by the spatial filter (hereinafter referred to as “size parameter”) and the weight of the pixel signal (hereinafter referred to as “intensity parameter”) are based on the pixel signal of each pixel of interest. It is determined. FIG. 4 is a flowchart showing an example of the operation flow of the noise removal circuit 205C that determines the size parameter and the intensity parameter of the spatial filter.

  When noise removal processing by the noise removal circuit 205C is started (S101), the pixel signals of all the pixels are separated into color components and luminance components (S102).

  In the processing step S103, an average value (average value of all pixels) is calculated for one of the separated color components (here, R component). When the average value of the R components of all the pixels is calculated, a size parameter corresponding to the average value is selected from a plurality of size parameters stored in advance in a predetermined storage area (S104). In the present embodiment, one size parameter is selected from the two size parameters depending on whether or not the average value of the R component is larger than a predetermined first threshold value.

  The selection of the size parameter will be specifically described. As a premise, the pixel signal includes a signal component having a level corresponding to the received light amount and a noise component superimposed on the signal component. This type of noise component includes, for example, random noise generated by an amplifier mounted on the image sensor 102, shot noise caused by dark current, and the like. When the average value of the R component is larger than the predetermined first threshold value, the level of the R component is likely to be relatively high with respect to the noise component, so that the SN ratio of the R component is estimated to be high. On the other hand, when the average value of the R component is equal to or less than the predetermined first threshold, the ratio of the noise component to the R component is likely to be higher (than the former), and therefore the SN ratio of the R component (compared to the former case) ) Estimated low.

  Also, in the spatial filter processing, the larger the size parameter, the higher the effect of improving the SN ratio because the number of pixels from which the weighted average value is taken increases. However, as the size parameter increases, the edge of the subject image based on the pixel signal subjected to the spatial filter processing becomes blurred.

  Therefore, in processing step S104, when the average value of the R components is larger than the predetermined first threshold, a small size parameter is selected. That is, for a pixel that is estimated to have a high S / N ratio, a size parameter suitable for suppressing a reduction in the edge of the subject image is selected. When the average value of the R component is equal to or smaller than the predetermined first threshold, a large size parameter is selected. That is, a size parameter with a high effect of improving the S / N ratio is selected for a pixel estimated to have a low S / N ratio.

  When a small size parameter is selected, for example, spatial filter processing is performed on the pixel signal of the target pixel based on the pixel signals of nine pixels of 3 × 3 centering on the target pixel. When a large size parameter is selected, for example, a spatial filter process is performed on the pixel signal of the target pixel based on the pixel signals of 25 pixels of 5 × 5 centered on the target pixel.

  When the size parameter of the spatial filter for the R component of each pixel is selected in processing step S104, it is determined whether or not the size parameter of the spatial filter is selected for all color components (S105). If there remains a color component for which no size parameter has been selected (S105: NO), the process returns to step S103. Here, after the R component, the processing steps S103 and S104 are sequentially executed for the remaining color components (B component and G component).

  As described above, in the endoscope system 1 of the present embodiment, the size parameter of the spatial filter processing of each target pixel is determined based on the average value of the color components of all the pixels, not for each pixel. Here, the main observation target of the endoscope system 1 is in a human lumen. Since the color of the human lumen is predominantly red, which is the color of blood, the pixel signal has a relatively large proportion of the R component in any pixel of interest. As a matter of course, the average value of the R components of all the pixels is also relatively larger than the other color components. Therefore, even if the size parameter of the spatial filter processing of each pixel is determined based on the average value of all the pixels, a size parameter suitable for each color component is selected. Thus, by determining the size parameter based on the average value of the color components of all pixels, the calculation load for determining the size parameter can be reduced.

  If the size parameter of the spatial filter is selected for all color components (S105: YES), the process of this flowchart proceeds to process step S106. In the processing step S106, for one pixel among the plurality of pixels, the difference between the luminance component of the pixel signal and one of the plurality of color components (here, R component) (hereinafter referred to as “R component difference”). ").) Is calculated. When the R component difference is calculated, an intensity parameter corresponding to the calculated R component difference is selected from a plurality of intensity parameters stored in advance in a predetermined storage area (S107). In the present embodiment, one intensity parameter is selected from two intensity parameters depending on whether or not the R component difference is larger than a predetermined second threshold. Note that the two intensity parameters that are selection candidates are appropriately changed according to the size parameter selected in processing step S104.

  The intensity parameter indicates the weight of each pixel signal of the pixel group (target pixel and its surrounding pixels) defined by the size parameter. The smaller the intensity parameter, the greater the weight of the pixel signal of the pixel of interest, and the greater the weight of the peripheral pixels closer to the pixel of interest. On the other hand, as the intensity parameter increases, the weight of the pixel signal of the target pixel decreases and the weight of the surrounding pixels far from the target pixel increases. Thereby, the difference in weight between the pixel signals of each pixel is reduced. In the spatial filter processing, the larger the intensity parameter, the higher the interpolation effect by surrounding pixels, and the higher the S / N ratio improvement effect. However, as the intensity parameter is larger, the edge of the subject image based on the pixel signal subjected to the spatial filter processing becomes blurred.

  Here, the luminance component is obtained by adding each color component at a predetermined ratio. Therefore, when the R component difference is large, the ratio of the R component in the pixel signal is low. Therefore, in the processing step S107, when the R component difference is larger than the predetermined second threshold, a large intensity parameter is selected. That is, for the R component having a low ratio in the pixel signal, an intensity parameter having a high effect of improving the SN ratio is selected. At this time, since the ratio of the R component is low, even if the edge of the subject image based on the R component is unclear, the influence on the sharpness of the edge of the subject image based on all the R, G, and B pixel signals is not affected. small.

  On the other hand, when the R component difference is small, the ratio of the R component in the pixel signal is high. Therefore, in the processing step S107, when the R component difference is equal to or smaller than the predetermined second threshold, a small intensity parameter is selected. That is, for the R component having a high ratio in the pixel signal, an intensity parameter suitable for suppressing a reduction in the sharpness of the edge of the subject image is selected.

  Fig.5 (a)-FIG.5 (d) are the figures for demonstrating the specific example of a spatial filter. 5A to 5D show a spatial filter for the target pixel P2 surrounded by a thick solid line. FIG. 5A shows a spatial filter when the size parameter is small and the intensity parameter is small. FIG. 5B shows a spatial filter when the size parameter is small and the intensity parameter is large. FIG. 5C shows a spatial filter when the size parameter is large and the intensity parameter is small. FIG. 5D shows a spatial filter when the size parameter is large and the intensity parameter is large.

  As shown in FIGS. 5A to 5D, the larger the size parameter, the greater the number of pixels for which the weighted average value of the pixel signal is calculated. Further, as the intensity parameter is larger, the weight of the pixel signal of the target pixel P2 is smaller, and the weight of the pixel signal of the surrounding pixels is relatively larger. In addition, the larger the size parameter or the larger the intensity parameter, the higher the interpolation effect by surrounding pixels, so the SN ratio is improved. In addition, the weight of the pixel signal of each pixel shown by Fig.5 (a)-FIG.5 (d) is an example, and is not limited to this. For example, the weight of the pixel signal of each pixel may be determined using a known Gaussian filter.

  When the intensity parameter of the spatial filter for the R component is selected in process step S107, it is determined whether or not the intensity parameter of the spatial filter is selected for all color components (S108). When the color component for which the intensity parameter is not selected remains (S108: NO), the process returns to the processing step S106. Here, the processing steps S106 and S107 are sequentially executed for the remaining color components (B component and G component) after the R component. When the intensity parameter of the spatial filter is selected for all color components (S108: YES), it is determined whether or not the intensity parameter has been selected for all pixels (S109).

  If there remains a pixel for which the intensity parameter is not selected (S109: NO), the process returns to the processing step S106, and the processing steps S106 and S107 are executed for the next pixel. On the other hand, when the intensity parameter is selected for all the pixels (S109: YES), a space using the selected spatial filter (size parameter and intensity parameter) for the R component, G component, and B component of each pixel. Filter processing is performed (S110). When the spatial filter processing is performed, the noise removal processing (SN ratio improvement processing) by the noise removal circuit 205C ends.

  As described above, in the present embodiment, the size parameter and the intensity parameter of the spatial filter for each pixel are determined (selected) based only on the pixel signal of the pixel and the average value of the pixel signals of all the pixels.

  When observing the inside of a human lumen using the endoscope system 1 of the present embodiment, the average value of the R component among the color components of the pixel signal is likely to be larger than the predetermined first threshold, and the B component. And the average value of the G component tends to be equal to or lower than the predetermined first threshold value. In addition, the R component difference tends to be less than or equal to the predetermined second threshold, and the B component difference and the G component difference tend to be larger than the predetermined second threshold. Therefore, a small size parameter and a small intensity parameter are selected for the R component, and a reduction in the sharpness of the edge of the subject image based on the R component of the pixel signal due to the noise removal processing is suppressed. Also, a large size parameter and a large intensity parameter are selected for the B component and the G component, and the SN ratio of the subject image based on the B component and the G component of the pixel signal is improved by the noise removal process. As a result, the pixel signal that has been subjected to the noise removal process maintains the sharpness of the edge of the dominant color (red in the case of the body cavity) and the SN ratio is improved.

  In addition, the calculation of the average value of all pixels has a small calculation load because the calculation method is simple. Further, in the present embodiment, unlike the conventional technique, the component separation process into the skeleton component and the residual component of the pixel signal having a relatively large calculation load and the correlation process between each pixel and the peripheral pixel are not performed. Therefore, according to the present embodiment, it is possible to perform noise reduction processing with few hardware resources on the pixel signal.

  The above is the description of the embodiment of the present invention. The present invention is not limited to the above-described configuration, and various modifications can be made within the scope of the technical idea of the present invention.

  For example, in this embodiment, each of the size parameter and the intensity parameter is selected from one of two patterns as shown in FIGS. 5A to 5B, but the present invention is not limited to this. Not. Each of the size parameter and the intensity parameter may be selected from three or more patterns.

  In this embodiment, the size parameter is selected based on the average value of the pixel signal, and the intensity parameter is selected based on the difference between the luminance component and the color component of the pixel signal. However, the present invention is not limited to this. . The size parameter may be selected based on the difference between the luminance component and the color component of the pixel signal. The intensity parameter may be selected based on the average value of the pixel signals. Further, the size parameter may be constant, and only the intensity parameter may be selected from a plurality of intensity parameters. Further, the intensity parameter may be constant and only the size parameter may be selected from a plurality of size parameters.

  In the present embodiment, the size parameter and the intensity parameter are stored in advance, but the present invention is not limited to this. The size parameter may be calculated by calculation using an average value of pixel signals. The intensity parameter may be calculated by calculation using a difference between the luminance component and the color component of the pixel signal.

  Further, in the processing step S102 of the flowchart shown in FIG. 4, the pixel signal is separated into a color component and a luminance component, but the present invention is not limited to this. For example, the pixel signal may be separated into a color difference component and a luminance component instead of being separated into a color component and a luminance component. Depending on the signal transmission method between the endoscope processor 200 and the monitor 300, the video signal may include a color difference signal and a luminance component. In this case, the image processing unit 205 uses the color difference in accordance with a predetermined color space (for example, YUV color space, YCbCr color space) before the noise removal processing by the noise removal circuit 205C. The signal and the luminance signal are separated. The size parameter and the intensity parameter of the spatial filter may be determined using the color difference component and the luminance component.

  The image sensor 102 includes R, G, and B pixels each having red (R), green (G), and blue (B) color filters, but the present invention is not limited to this. For example, the image sensor 102 may include G, Cy, Mg, and Ye pixels having green (G), cyan (Cy), magenta (Mg), and yellow (Ye) color filters, respectively. In this case, after demosaic processing is performed on each pixel signal, color conversion processing is performed to generate a pixel signal having R, G, and B color information.

DESCRIPTION OF SYMBOLS 1 Endoscope system 100 Electronic scope 101 Objective optical system 102 Image sensor 103 Image sensor driver 104 AFE (Analog Front End)
105 Illumination optical system 106 Light guide 200 Endoscope processor 201 System controller 202 Timing controller 203 Light source 204 Light source driver 205 Image processing unit 205A Demosaic processing circuit 205B Color adjustment circuit 205C Noise removal circuit 206 Image memory 207 Video signal generation circuit 208 Front panel 300 monitors

Claims (8)

An endoscope processor that processes a pixel signal of each pixel of a subject imaged by a predetermined imaging device,
Component separating means for separating a pixel signal of a target pixel in the image sensor into a luminance component and a color information component ;
Parameter determining means for determining a parameter of a predetermined spatial filter related to the target pixel based on at least one of the separated luminance component and color information component ;
By performing a predetermined spatial filtering on the pixel of interest based on a parameter of a predetermined spatial filter the determined comprises a noise reduction means for reducing the noise of the target pixel, and
The noise reduction means includes
Based on the parameter of the predetermined spatial filter determined by the parameter determining means, the weighted average of pixel signals is calculated for each color information component for the target pixel and peripheral pixels around the target pixel. Apply spatial filtering to the pixel signal of the pixel
Endoscopic processor . The parameter of the predetermined spatial filter is:
A size parameter indicating the number of the surrounding pixels from which the weighted average of the pixel signals is taken and the position of the surrounding pixels relative to the pixel of interest;
An intensity parameter indicating the weight of each pixel signal for which the weighted average is taken;
The endoscope processor according to claim 1 . The parameter determination means includes
For each color information component , calculate an average value of all the pixel signals, and determine a parameter of the predetermined spatial filter based on the calculated average value.
The endoscope processor according to claim 1 or 2 . The parameter determination means includes
For each pixel, calculate the difference between the luminance component and the color information component, and determine the parameters of the predetermined spatial filter based on the calculated difference.
The endoscope processor according to any one of claims 1 to 3 . The parameter determination means includes
For each color information component , calculate an average value of all the pixel signals, and determine one of the size parameter and the intensity parameter based only on the calculated average value.
The endoscope processor according to claim 2 . The parameter determination means includes
For each pixel, calculate the difference between the luminance component and the color information component, and determine the other of the size parameter and the intensity parameter based only on the calculated difference.
The endoscope processor according to claim 5 .   An endoscope processor that processes a pixel signal of each pixel of a subject imaged by a predetermined imaging device,
  Component separating means for separating a pixel signal of a target pixel in the image sensor into a luminance component and a color information component;
  Parameter determining means for determining a parameter of a predetermined spatial filter related to the pixel of interest based on the separated luminance component and color information component;
  Noise reduction means for reducing noise of the target pixel by performing predetermined spatial filter processing on the target pixel based on the determined parameter of the predetermined spatial filter;
Comprising
Endoscopic processor. The endoscope processor according to any one of claims 1 to 7,
An electronic scope having the predetermined imaging device,
Endoscope system.

Images