JPWO2018079071A1 - Imaging device - Google Patents

Abstract

Abstract: A technique is provided for reducing noise and reduction in spatial resolution in a mixing and adding process in an imaging apparatus and alleviating a video shock that occurs at the time of process switching. The pixel mixing unit 100a of the imaging apparatus includes a pixel extraction circuit 300, a parallel subblock group 301 provided in parallel, a selector group 302 that controls the output of each subblock, and a gain control unit 304 that controls the selector group 302. And a mixing and adding circuit 303 for adding the output of the parallel sub-block group 301. The parallel sub block group 301 includes sub blocks (1) 301_1 to sub blocks (12) 301_12 which are twelve sub blocks. Here, as pixel values to be added to the target pixel, predetermined four pixels equidistant from the target pixel are set as one group, and an average value of pixel values of four pixels is used for pixel values in the same group. Then, pixel values of grouped pixels are input to each sub block.

Description

  The present invention relates to an imaging device having a solid-state imaging device.

  In recent years, awareness of “safety” and “reliability” has increased, and the demand for imaging devices for surveillance such as home security is growing. The performance of the imaging device depends on high definition, wide dynamic range and high sensitivity. In particular, photographing under a low illumination environment is frequently performed not only for surveillance imaging devices but also for consumer imaging devices used in home video for home use, etc., and imaging devices having high sensitivity Is required. In these imaging devices, in order to increase the imaging sensitivity even in a low illuminance environment, a sensitivity increase method in which the shutter speed of the imaging device is controlled to perform exposure for a long time, and pixel values of a plurality of adjacent pixels in the imaging device A method called binning (binning process) is used in which the data is collectively read out collectively. In addition, in video processing after being read out from an imaging device, a method called digital pixel addition in which adjacent pixels are mixed and added is also used (see, for example, Patent Document 1).

  By the way, in the pixel mixing method in the conventional imaging device, improvement is still required regarding noise and resolution. Generally, in order to obtain high imaging sensitivity, it is necessary to increase the amount of light received by the imaging device. However, the multi-pixel image pickup device employed in an image pickup apparatus in pursuit of high definition has a narrow light receiving area per pixel due to the limitation of the element size, and it is difficult to further increase the sensitivity. . For this reason, in recent imaging apparatuses, a method (known method) such as digital pixel addition is used in which pixels adjacent in space are mixed in signal processing to increase the gain with respect to pixel values read out from the imaging element.

JP, 2013-110644, A

  However, the above-described known method has a problem that as the number of pixels to be added is increased to increase the sensitivity, noise (quantization noise) is also increased, and the spatial resolution is lowered. Furthermore, in a general imaging device, the processing is switched in two steps of ON or OFF, so there is a problem that an unnatural video shock appears when switching.

  The present invention has been made in view of such a situation, and an object thereof is to solve the above-mentioned problems.

The present invention is an imaging device comprising: a solid-state imaging device; and a pixel mixing unit that adds the pixel values of peripheral pixels to the pixel value of a pixel of interest to increase the sensitivity of the solid-state imaging device, The mixing unit treats as a group a total of four pixels including two pixels which are symmetrical with respect to the set of two pixels symmetrical with respect to the target pixel as one set, and the peripheral pixels are set to the pixel value of the target pixel. When adding the pixel values of (1), the average value of the pixel values of the same group is used as the pixel value of the pixels belonging to the same group, and when mixing the pixel values of pixels having high correlation with the target pixel. A large sensitivity is reflected, and when mixing pixel values of low correlation pixels, a small sensitivity is reflected and added.
In addition, the pixel mixing unit includes a plurality of sub blocks provided in parallel in group units, a plurality of selectors controlling outputs of the respective sub blocks, a gain control unit controlling the selectors, and the plurality of sub blocks. And an adder circuit for adding the outputs of the subblocks.
The pixel mixing unit may be provided inside the solid-state imaging device.

  According to the present invention, it is possible to provide a sensitivity-up method for reducing the noise and the reduction of spatial resolution that occur in the mixing and adding process, and further alleviating the video shock that occurs when switching the process.

It is a functional block diagram of a 3 board type imaging device concerning a 1st embodiment. It is a block diagram of a pixel mixing part of a solid-state image sensing device concerning a 1st embodiment. FIG. 6 is a diagram showing the relationship between sub-blocks and extracted pixels to be the input according to the first embodiment. It is an example of the conversion table with which the gain control part based on 1st Embodiment is provided. It is a block diagram of the 3 board type imaging device of a comparative example concerning a 1st embodiment. It is a functional block diagram of a single type imaging device concerning a 2nd embodiment. It is a functional block diagram of the single type imaging device of a comparative example concerning a 2nd embodiment.

  Next, modes for carrying out the present invention (hereinafter, simply referred to as “embodiments”) will be specifically described with reference to the drawings. In the present embodiment, pixel addition processing is performed using pixels in the N × N area around the pixel of interest. At that time, the pixel area size is adaptively switched by predetermined processing according to the sensitivity increase setting, and in accordance with the sensitivity increase degree, generation of noise is suppressed and reduction in spatial resolution is reduced. Furthermore, a plurality of pixels located point-symmetrically to the target pixel are grouped, the pixel average values of the respective groups are combined, mixed addition processing is performed on an apparent one-pixel basis, and an image by level difference that may occur at sensitivity switching. Reduce the shock.

  Hereinafter, in the first embodiment, the three-plate type imaging device 20a will be specifically described, and in the second embodiment, the single-plate type imaging device 20b will be specifically described.

  First Embodiment FIG. 1 is a functional block diagram of a three-plate type imaging device 20a according to the present embodiment. The three-plate type imaging device 20a mainly includes a lens 201, a spectral filter 202, a solid-state imaging device 203, an image processing unit 207, and an image output unit 211.

  A subject image collected by the lens 201 is split into red (R) component, green (G) component and blue (B) component by the spectral filter 202 and taken into three solid-state imaging devices 203. As described above, the color filter array (CFA) is not provided in the solid-state imaging device 203 in order to disperse light with the spectral filter 202.

  The solid-state imaging device 203 is, for example, a semiconductor image sensor such as a CCD image sensor or a COMS image sensor. The solid-state imaging device 203 includes a solid-state imaging device (R) 203R, a solid-state imaging device (G) 203G, and a solid-state imaging device corresponding to each of red (R) component, green (G) component, and blue (B) component light. And an element (B) 203B.

  That is, the light of the red (R) component is taken into the solid-state imaging device (R) 203R. The light of the green (G) component is taken into the solid-state imaging device (G) 203G. The light of the blue (B) component is taken into the solid-state imaging device (B) 203B. The solid-state imaging device (R) 203R, the solid-state imaging device (G) 203G, and the solid-state imaging device (B) 203B are elements having the same structure, and unless otherwise distinguished, they are simply referred to as "solid-state imaging device 203". It is called and explained.

  Each solid-state imaging device 203 includes an AMP (amplifier unit) 204, an ADC (A / D converter) 205, a pixel mixing unit 100a, and an IF unit 206.

  The AMP 204 multiplies the signal of the acquired component, and executes the above-mentioned binning process if necessary. For example, the AMP 204 of the solid-state imaging device (B) 203B electrically amplifies a signal in which light of blue (B) component is photoelectrically converted in each pixel, and simultaneously adds signals (charges) of adjacent pixels Execute the binning process.

  The ADC 205 converts the multiplied signal into a digital value of 8 bits to 14 bits per pixel, and outputs the digital value to the pixel mixing unit 100 a as a pixel value (also referred to as “pixel data”).

  The pixel mixing unit 100 a performs signal processing using a pixel mixing method described later, and outputs the signal processing to the IF unit 206. The IF unit 206 adds a synchronization word consisting of a unique data string for frame synchronization to the pixel data output from the pixel mixing unit 100a, and then transfers the pixel data to the video processing unit 207.

  Similar to the solid-state imaging device 203, the image processing unit 207 is an image processing unit (R) 207R corresponding to each of red (R) component, green (G) component, and blue (B) component light. G) 207G, and an image processing unit (B) 207B. That is, the output of the solid-state imaging device (R) 203R is taken into the video processing unit (R) 207R. The output of the solid-state imaging device (G) 203G is taken into the video processing unit (G) 207G. The output of the solid-state imaging device (B) 203B is taken into the video processing unit (B) 207B. Here, the video processing unit (R) 207R, the video processing unit (G) 207G, and the video processing unit (B) 207B have a common structure, and in the following, simply “video processing unit” unless otherwise specified. This will be described as “207”.

  The video processing unit 207 includes an IF unit 210, a signal processing unit 209, and a pixel mixing unit 100b.

  The IF unit 210 acquires pixel data from the solid-state imaging device 203, detects a synchronization word added to the pixel data, and generates a video timing signal as a reference of post-stage processing. That is, the subsequent processing is synchronized with the timing signal generated by the IF unit 210.

  The signal processing unit 209 improves image processing such as correction processing such as scratches (pixel defects), gamma correction and knee correction for securing a dynamic range, and visibility of pixel data supplied from the IF unit 210. To perform signal processing such as noise reduction.

  After the general processing, the pixel mixing unit 100b performs pixel mixing processing to increase sensitivity. Here, the same processing as the pixel mixing unit 100 a of the solid-state imaging device 203 is performed. The output of the ADC 205 in the solid-state imaging device 203 described above is digitized red (R), green (G), or blue (B) pixel data. That is, since the pixel mixing unit 100a in the solid-state imaging device 203 has the same interface as the pixel mixing unit 100b in the video processing unit 207, the same pixel mixing process can be efficiently performed in the two pixel mixing units 100a and 100b. .

  The pixel data of each light processed by the three video processing units 207 is converted into various standard formats by the video output unit 211 and is output as a video signal.

  Subsequently, the pixel mixing process will be described with reference to FIGS. FIG. 2 is a block diagram of the pixel mixing unit 100 a of the solid-state imaging device 203. Here, the mixing process of 7 × 7 pixels at maximum is described as an example. The pixel mixing unit 100 b of the video processing unit 207 also has the same function and configuration, and thus the description thereof is omitted.

  The mixed pixel unit 100a receives pixel data and a 5-bit gain setting parameter signal (Gain_Para [4: 0]), and outputs mixed and added pixels mixed and added in this block. Specifically, the pixel mixing unit 100a includes a pixel extraction circuit 300, a parallel subblock group 301, selector groups 302 (302_1 to 302_12), a mixing and adding circuit 303, and a gain control unit 304.

  The parallel sub-block group 301 is composed of sub-blocks (1) 301_1 to sub-blocks (12) 301_12 which are twelve sub-blocks provided in parallel.

  The selector group 302 includes twelve selectors SEL (1) 302_1 to SEL (12) 302_12. The SEL (1) 302_1 to the SEL (12) 302_12 switch the outputs of the sub block (1) 301_1 to the sub block (12) 301_12, respectively.

  The gain control unit 304 applies a 6-bit control signal (Gain [5: 0]) for controlling the selector group 302 (sub block (1) 301_1 to sub block (12) 301_12) to control them. . The correspondence between Gain [5: 0] and Gain_Para [4: 0] is preset.

  The pixel extraction circuit 300 delays input pixel data using a line memory and a delay element, and extracts a pixel of interest (center of gravity pixel) Pix — 33 and peripheral pixels located in the vicinity thereof. In the present embodiment, 49 pixels corresponding to 7 × 7 pixels are simultaneously output by the pixel extraction circuit 300.

  Also, the 49 pixels output from the pixel extraction circuit 300 are distributed to the parallel sub-block group 301 (sub-blocks (1) 301_1 to sub-blocks (12) 301_12) at the subsequent stage by the positional relationship with the target pixel Pix_33. Do the processing. Note that the configuration shown in FIG. 2 is for explaining the principle, and in practice, a method (Maclelan transformation or the like) for performing the equivalent operation more efficiently may be implemented.

  FIG. 3 is a diagram showing the relationship between the subblocks and the extracted pixels to be the input. Specifically, FIGS. 3A to 3L correspond to sub block (1) 301_1 to sub block (12) 301_12, respectively.

  A value in units of 4 pixels (that is, 4 pixel values), in which a set of two pixels located point-symmetrically with respect to a pixel of interest Pix_33 serving as the central pixel is one set, and four sets of the set are vertically or horizontally symmetrical The addition process is performed using a value based on the average of From the total gain of 4 pixels, gains of 1⁄4 (equivalent to 1 pixel), 2/4 (equivalent to 2 pixels), 3⁄4 (equivalent to 3 pixels) are calculated.

  For example, FIG. 3A shows four pixels to be input to sub block (1) 301_1, and four pixels (Pix_23, Pix_43, Pix_32, Pix_34) adjacent to upper, lower, right and left with respect to the pixel of interest Pix_33 are subblocks. (1) It becomes an input pixel to 301_1.

  Similarly, in FIG. 3B, the pixel input to the sub block (2) 301_2 is four pixels (Pix_22, Pix_44, Pix_24, Pix_42) diagonally adjacent to the pixel of interest (Pix_33).

  Thus, the smaller the sub-block number, the closer the distance from the target pixel Pix_33, and the higher the correlation with the target pixel Pix_33. Conversely, the larger the sub-block number, the farther the distance from the target pixel Pix_33. It means that it is a block with low correlation.

  In this example, when the outputs of the subblocks are added, high sensitivity (gain) is set for an output having a small subblock number, and 0 or small sensitivity (gain) is set for an output having a large subblock number. This can be considered as an approximation of FIR filter coefficients represented by a sinc function or a Gaussian function.

  Also, in order to obtain the gain to be set, sub-blocks with the smallest sub-block number are sequentially added. For example, when gain corresponding to 9 pixel addition is obtained, the total gain of each four pixels for sub block (1) 301_1 and sub block (2) 301_2 is the total of four pixels for sub block (3) 301_3. One fourth of the gain (corresponding to one pixel) is used for the addition. Here, the sub block (4) 301_4 is not used in place of the sub block (3) 301_3. That is, when the pixel area size (area of the pixel to be mixed) is adaptively switched by predetermined processing according to the sensitivity increase setting, pixels of high correlation are sequentially added to the target pixel and priority is given to the pixel area to be switched. Can be done. Generally, pixels in areas closer in space show higher correlation to the target pixel, but it can be assumed that spatial distance and actual correlation may differ depending on the shooting target, shooting time, etc. In this case, subblocks may be replaced as appropriate. Subblock (1) 301_1 and subblock (2) 301_2, subblock (3) 301_3 and subblock (4) 301_4, etc., which have the same spatial distance, are used in adaptive color plane interpolation (ACPI) etc. In the same manner as in the method described above, the order of addition can be adaptively switched.

  Each sub block (sub block (1) 301 _ 1 to sub block (12) 301 _ 12) includes an adder circuit 306 and a multiplier circuit 307. The addition circuit 306 performs addition processing of four input pixels. The multiplication circuit 307 outputs an operation value obtained by multiplying the output value of the addition circuit 306 by 1/4, 2/4, and 3/4 coefficients. These operation values are mixed addition values of one pixel, two pixels, three pixels, and four pixels in each sub block (sub block (1) 301_1 to sub block (12) 301_12).

  The selector group 302 (SEL (1) 302_1 to SEL (12) 302 12) corresponds to one pixel, two pixels, three pixels and four pixels in each sub block (sub block (1) 301_1 to sub block (12) 301_12) The selected Sub_out_1 to Sub_out_12 signals are input to the final stage of the mixing and adding circuit 303.

  Next, 6-bit control signals for controlling the selector group 302 (SEL (1) 302_1 to SEL (12) 302_12) will be described. The 6-bit control signal (Gain [5: 0]) is output from the gain control unit 304.

  FIG. 4 shows an example of the conversion table provided in the gain control unit 304. The gain control unit 304 sets a 5-bit gain setting parameter (Gain_Para [4: 0]), which means the corresponding pixel mixing number, based on this conversion table, to uniquely output a 6-bit output control signal (Gain). [5: 0]), ie, output the set gain value.

  Of the 6 bits, a sub block to which 4 bits on the MSB (Most Significant Bit) side are applied, and 2 bits on the LSB (Least Significant Bit) side are allocated to the output selection of the sub block.

  For example, when Gain_Para [4: 0] = 00000b in which the number of mixed pixels means 0 is input, the 6-bit control signal (Gain [5: 0]) output according to the conversion table of FIG. 4 becomes 00h. As a result, all the selector groups 302 (SEL (1) 302_1 to SEL (12) 302_12) in FIG. 4 are turned off, and the operation is as expected.

  In addition, when setting (Gain_Para [4: 0] = 01010b) to make the number of pixels to be mixed equal to 10, Gain [5: 0] = 0Dh is obtained through the conversion table of FIG. 1) 301_1, sub-block (2) 301_2 and sub-block (3) 301_3 are selected, and in the final stage mixing and adding circuit 303, mixing processing of pixels is performed using 12 pixels relating to the above-mentioned sub block, 10 pixels It is output as a considerable addition pixel. That is, the number of pixels to be subjected to pixel addition is 4n + 1 (n is a natural number). In gain setting equivalent to 10 to 13 pixels, internally, 13 pixels obtained by combining n subblocks and the central pixel are used for the mixing process.

  In the conversion table in the figure, when viewed in units of subblocks, the relationship between the number of pixels to be mixed and the gain is, for example, a relationship corresponding to a logarithmic function. The sub-block unit is because an average of four pixel values is used in the step in the same sub-block, so that a slight blurring with the logarithmic function occurs strictly in the same sub-block.

  According to the above processing, the pixel area size can be adaptively switched by predetermined processing according to the sensitivity increase setting, and according to the sensitivity increase degree, generation of noise can be suppressed and decrease in spatial resolution can be reduced. it can.

  Furthermore, in the proposed method, a plurality of pixels located point-symmetrically with respect to the corresponding pixel are grouped, the pixel average values of the respective groups are combined, and mixed addition processing is performed on an apparent one-pixel basis. It is possible to reduce the video shock due to the level difference.

  FIG. 5 is a block diagram of a three-plate type imaging device 200a of a comparative example. The difference between the three-plate type imaging device 20a shown in FIG. 1 and the three-plate type imaging device 200a in FIG. 5 is that the pixel mixing unit 100a of the solid-state imaging device 203 is provided in the three-plate type imaging device 20a of FIG. The point 5 is that the pixel mixing unit is not provided. Furthermore, the pixel mixing unit 208 of the three-plate type imaging device 200a of FIG. 5 performs pixel mixing processing for increasing sensitivity. In the pixel mixing unit 208, for example, the pixel values of two pixels spatially adjacent in the horizontal direction are mixed and added, and the sensitivity is doubled. Due to such processing, the pixel area size can not be adaptively changed by predetermined processing according to the sensitivity increase setting. In addition, it is not possible to prevent video shock due to a level difference that may occur at the time of sensitivity switching.

  Note that only one of the pixel mixing unit 100 a of the solid-state imaging device 203 and the pixel mixing unit 100 b of the video processing unit 207 may be used as the configuration for performing the pixel mixing process. In addition, the pixel mixing unit 100a of the solid-state imaging device 203 may perform the process proposed in the present embodiment, and the video processing unit 207 may perform the process of the pixel mixing unit 208 shown in the comparative example. Also, the pixel mixing performed by the pixel mixing units 100a and 100b can be performed in cooperation with binning performed in the element. In a simple example, if 2 × 2 or 4 × 4 pixels can be optically combined and generated as one pixel by the binning processing mode, the 4 or 16 times gain increase is not in the pixel mixing unit 100 a or the like. This is performed by binning, and the gain of fine steps other than that is performed by the pixel mixing unit 100a or the like. Assume that the target pixel mixture range (7 × 7, 5 × 5, 3 × 3) is switched. At that time, with respect to subblocks within the respective mixed ranges, weighting is appropriately changed, and mixed addition is performed.

  Second Embodiment FIG. 6 is a block diagram of a single-plate type imaging device 20b according to the present embodiment. In this embodiment, the pixel mixing process of the first embodiment (three-plate type imaging device 20a) is applied to a single-plate type imaging device 20b. About the structure of the same function, the same code | symbol is attached | subjected and description is abbreviate | omitted suitably.

  The single-plate type imaging device 20 b includes a lens 201, an imaging element 213, an image processing unit 214, and an image output unit 211.

  Instead of the spectral filter 202 included in the three-plate type imaging device 20 a of the first embodiment, an imaging element 213 having an on-chip CFA 212 is used.

  An object image collected by the lens 201 becomes color information of red (R), green (G) and blue (B) by the CFA 212 formed inside the imaging device 213, and is sent to the AMP 204.

  As described above, in the three-plate type imaging device 20a according to the first embodiment, amplification of red (R), green (G), and blue (B) components is performed by the AMPs 204 respectively included in the three imaging elements 203. In the single-plate type imaging device 20b, since each pixel has only single color information of red (R), green (G) and blue (B), a single pixel can be applied by applying different pixel gains to each color. Amplification and the like can be performed with one AMP 204.

  The ADC 205 converts the color information of red (R), green (G) and blue (B) acquired from the AMP 204 into digitized pixel data, and outputs a Bayer array signal to the pixel mixing unit 101.

  The pixel mixing unit 101 performs the same image mixing process for each color as the pixel mixing unit 100 a of the first embodiment, and outputs the image mixing process to the IF unit 206. As a method of processing for each color, there is a method of providing and switching the conversion table of FIG. 4 for each color, and a method of separating the signal for each color and providing the configuration of the pixel mixing unit 100a of FIG. The IF unit 206 adds a unique synchronization word to pixel data, and transfers the pixel data to the video processing unit 214.

  In the video processing unit 214, the IF unit 210 performs synchronous word detection similar to that of the three-chip imaging device 200a, and generates a video timing signal as a reference. However, at this time, since each pixel data of red (R), green (G) and blue (B) has the above-mentioned Bayer arrangement, it is arranged only at every other pixel. Therefore, the Bayer conversion unit 215 in the subsequent stage performs demosaicing processing and complements the non-existent red (R), green (G), and blue (B) color information based on the information of the peripheral pixels.

  Therefore, in the subsequent processing of the Bayer conversion unit 215, each pixel has color information of red (R), green (G), and blue (B), respectively, and each is processed independently. That is, the signal processing unit 209 includes a signal processing unit (R) 209R for red (R) color information, a signal processing unit (G) 209 G for green (G) color information, and a blue (B) color. A signal processing unit (B) 209B for information is provided.

  The pixel mixing unit 100 further includes a pixel mixing unit (R) 100R, a pixel mixing unit (G) 100G, and a pixel mixing unit (B) 100B. The pixel mixing unit (R) 100R performs pixel mixing processing on the signal acquired from the signal processing unit (R) 209R. The pixel mixing unit (G) 100G performs pixel mixing processing on the signal acquired from the signal processing unit (G) 209G. The pixel mixing unit (B) 100B performs pixel mixing processing on the signal acquired from the signal processing unit (B) 209B.

  After the processing of the pixel mixing unit 100, pixel data of each color is converted into various standard formats by the video output unit 211, and is output as a video signal. The pixel arrangement in the single plate system is not limited to the Bayer system, and an arbitrary arrangement may be used. In that case, pixel extraction for each subblock as shown in FIG. 3 needs to be designed to match the symmetry in the pixel array of the imaging device.

  Here, FIG. 7 shows a single-panel type imaging device 200b of a comparative example. In the single-plate type imaging device 200b, the imaging device 213 is not provided with a pixel mixing function. Further, the processing of the pixel mixing unit 208 of the image processing unit 214 is, for example, pixel values of two pixels spatially adjacent in the horizontal direction, as in the three-plate type imaging device 200a of the comparative example of the first embodiment. Is mixed and added, and the sensitivity is doubled. Due to such processing, the pixel area size can not be adaptively changed by predetermined processing according to the sensitivity increase setting. In addition, it is not possible to prevent video shock due to a level difference that may occur at the time of sensitivity switching.

  On the other hand, in the single-plate type imaging device 20b of the present embodiment, the pixel area size can be adaptively switched by predetermined processing according to the sensitivity increase setting, and generation of noise can be suppressed according to the degree of sensitivity increase. The reduction in resolution can also be mitigated.

  Furthermore, in the proposed method, a plurality of pixels located point-symmetrically with respect to the corresponding pixel are grouped, the pixel average values of the respective groups are combined, and mixed addition processing is performed on an apparent one-pixel basis. It is possible to reduce the video shock due to the level difference.

  The present invention has been described above based on the embodiments. This embodiment is an exemplification, and various modifications and applications are possible for the combination of the respective constituent elements. For example, when using for compensation of a diffraction aberration etc., each gain setting or addition pattern with respect to pixel mixing part 100a, 100b of each color does not necessarily need to be the same. The configuration of the present embodiment is not limited to the one for the purpose of sensitivity enhancement, but can be operated as a spatial filter in a situation where sufficient illuminance or SNR is obtained, and compensation for the aperture characteristic of the imaging device It can be used for compensation of chromatic aberration, compensation of diffraction aberration called small aperture blur, and the like. For example, when using for compensation of a diffraction aberration etc., you may set up the addition pattern of the characteristic which compensates for the blurring which generate | occur | produces notably in B pixel according to a diaphragm value. Also, under ultra-low illumination environment, pixel mixing is performed only on R and G pixel data that have a large influence on visibility, and pixel mixing is not performed on B pixel data with relatively low sensitivity, and a useless resolution is returned. The decrease may be suppressed. Conversely, the pixel depth may be increased by performing pixel addition only for B pixel data, and color reproducibility may be enhanced while preventing SNR degradation due to linear matrix calculation. Thus, configurations of embodiments of the present invention may provide SNR-spatial resolution scalability in various ways.

  The present invention is applicable to an imaging device having a solid-state imaging device, a monitoring camera system, and the like. This application claims the benefit of priority based on Japanese Patent Application No. 2016-212803 filed Oct. 31, 2016, the entire disclosure of which is incorporated herein by reference.

20a Three-plate imaging device 20b Single-plate imaging device 100, 101, 100a, 100b Pixel mixing unit 100R Pixel mixing unit (R) 100G Pixel mixing unit (G) 100B Pixel mixing unit (B) 201 Lens 202 Spectral filter 203, 213 Solid Imaging device 203R Solid-state imaging device (R) 203G Solid-state imaging device (G) 203B Solid-state imaging device (B) 204 AMP 205 ADC 206, 210 IF unit 207, 214 Image processing unit 207R Image processing unit (R) 207G Image processing unit (G) 207B image processing unit (B) 209 signal processing unit 209R signal processing unit (R) 209G signal processing unit (G) 209B signal processing unit (B) 211 image output unit 212 CFA 215 Bayer conversion unit 300 pixel extraction circuit 301 parallel subblock group 301_1 to 301 12 sub-blocks (1) to the sub-block (12) 302 selector group 302_1~302_12 SEL (1) ~SEL (12) 303 mixed adder circuit 304 the gain control unit 306 adder circuit 307 multiplying circuit

Claims (3)

An imaging apparatus comprising: a solid-state imaging device; and a pixel mixing unit that adds the pixel values of peripheral pixels to the pixel value of a pixel of interest to increase the sensitivity of the solid-state imaging device,
The pixel mixing unit treats as a group a total of four pixels including two pixels symmetrical with respect to the set of two pixels symmetrical with respect to the set as a group, and sets the pixel value of the target pixel as a group. When adding the pixel values of peripheral pixels, using the average value of the pixel values of the same group as the pixel value of the pixels belonging to the same group and mixing the pixel values of pixels having high correlation with the target pixel An imaging apparatus characterized in that, when mixing pixel values of pixels with low correlation, the sensitivity is reflected and small sensitivity is reflected and added. The pixel mixing unit includes a plurality of sub-blocks provided in parallel in group units;
A plurality of selectors that control the output of each of the subblocks;
A gain control unit that controls the selector;
An adder circuit for adding the outputs of the plurality of subblocks;
The imaging device according to claim 1, comprising:   The imaging device according to claim 1, wherein the pixel mixing unit is provided inside the solid-state imaging device.

Images