JP2009130777A - Imaging apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve an image quality by making correction without affecting a normal video signal as a countermeasure to cope with a noise component that appears in a signal on the boundary of imaging areas. <P>SOLUTION: A read area front-part noise detection circuit 44 detects a high frequency component by calculating a difference between a level of a target pixel and a level average value of both neighboring pixels before and behind the target pixel for video signals within the range of several front-part pixels in a read area. As the high frequency component becomes larger, a possibility of being noise becomes higher. A read area rear-part high-luminance detection circuit 45 calculates an average value for video signals within the range of several rear-part pixels in the read area of a preceding line. A selector 4A mixes an interpolation signal created by an interpolation signal calculation circuit 41 and an input video signal stored in a memory 42 in a ratio corresponding to a correction coefficient α and a correction coefficient (1-α) from a multiplier 4F, thereby producing a corrected signal in which a noise component is reduced. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an imaging apparatus, and more particularly to an imaging apparatus using a divided readout imaging element.

  In recent years, high-pixel and high-definition imaging devices used in imaging devices are accelerating, and there are tens of millions of pixels for still images and more than 8 million pixels for moving images. Yes. With the increase in the number of pixels, particularly in the case of a moving image pickup element, in order to take out a picked-up signal at high speed, it is generally performed that the image pickup region is divided and divided reading is performed.

  This divided readout method has a problem that the characteristics of the sensitivity and output signal differ depending on the output amplifier and A / D converter used for each divided region of the image sensor, and various proposals have been made as methods for correcting this. (For example, refer to Patent Documents 1 and 2).

  Patent Document 1 examines an electrical signal in a specific range for each divided region, calculates a correction value for correcting a signal difference in a corresponding divided region based on the electric signal, and based on the correction value, A method is disclosed in which the gain and offset of an output amplifier or A / D converter used in a divided region including a specific range is adjusted to suppress a signal characteristic difference for each divided region. Further, in Patent Document 2, the gain and offset of an imaging signal transmission system in a divided area are calculated using an average value of signals having strong similarity at each boundary between divided areas and an average value of signals from adjacent divided areas. A method of adjusting and suppressing a signal characteristic difference for each divided region is disclosed.

JP 2003-018472 A JP 2004-179737 A

  However, the divided readout imaging device has other problems that cannot be solved by the inventions described in Patent Document 1 and Patent Document 2. That is, in the divided readout imaging device, noise in units of one pixel due to fluctuations in the power supply voltage due to a sudden change in the imaging signal or signal leakage due to routing of the signal wiring is present at the boundary between the imaging regions divided by the device. There is a problem that it may appear.

  For example, when a high-intensity signal is input to the last read pixel at the boundary of the divided area of a line and the range of the previous few pixels (hereinafter referred to as the rear pixel), The signal of the front pixel) may be affected, resulting in a signal obtained by adding a noise component to the normal signal. Since such noise components appear at the boundary of the divided areas, they often become the central part of the entire image, and are emphasized by subsequent gamma correction processing or the like. . However, at present, nothing is disclosed about countermeasures against such a phenomenon.

  The present invention has been made in view of the above points. As a countermeasure against noise components appearing in a signal at the boundary of an imaging region, the imaging capable of improving image quality by performing correction without affecting a normal video signal. An object is to provide an apparatus.

  In order to achieve the above-described object, the present invention has an image pickup apparatus of a divided read type having a plurality of read areas obtained by dividing an image pickup area of an image pickup device into a plurality, and reading video signals in parallel from each of the plurality of read areas. An interpolation signal calculation means for calculating an interpolation signal used for noise correction by linear interpolation from a video signal output from the readout area, a first pixel of a correction target line of the video signal output from the readout area, and In the first pixel range consisting of a predetermined number of pixels immediately after, the difference between the level of the target pixel and the level average value of each pixel adjacent to the target pixel is obtained, and the high frequency component is determined. A high-frequency component detecting means for detecting, and a video signal output from the readout area immediately before one line detected by the high-frequency component detecting means. Average value calculating means for calculating the average value of the video signal in the second pixel range consisting of the last read pixel at the boundary of the read area in the screen and a predetermined number of pixels immediately before it, and high frequency component detecting means Correction coefficient calculating means for calculating a correction coefficient by multiplying the high frequency component detected by the average value calculated by the average value calculating means, and a relative addition ratio of the interpolation signal to the video signal output from the readout area Is changed according to the value of the correction coefficient, and a circuit unit including an adding unit that generates a corrected signal by adding the video signal and the interpolation signal is provided for each of the video signals output from the plurality of readout regions. Is provided.

  In the present invention, when a high-intensity video signal is input to the second pixel range of a certain line, fluctuations in power supply voltage or signal wiring due to a sudden change in signal are input to the first pixel range of the next line. When noise due to signal leakage due to routing appears, the above average value becomes high, so that the value of the correction coefficient becomes high. As a result, the addition means performs processing on the pixel to be processed (input video signal). Thus, a relatively large amount of interpolation value components (interpolation signals) from both adjacent pixels are mixed, and the output corrected signal has a reduced noise component.

  Here, among the pixels of the second pixel range, the pixel signal value is a, the value of the pixel whose value is larger than a predetermined threshold value, and noise first appears in the second pixel range. Where b is the maximum value as the pixel signal value and c is a coefficient calculation means for calculating the coefficient represented by (ab) / (cb) instead of the average value calculation means. The calculating means may be means for calculating a correction coefficient by multiplying the coefficient calculated by the coefficient calculating means and the high frequency component detected by the high frequency component detecting means.

  According to the present invention, in an image pickup apparatus using a divided readout type image sensor, a high-luminance signal is generated in the second pixel range of the read area of the image sensor without affecting the image signal during normal imaging. When entering, the noise component added by the influence of the normal signal in the first pixel range of the next line is suppressed, and the image quality can be improved.

  Next, an embodiment of the present invention will be described in detail with reference to the drawings.

  FIG. 1 shows a block diagram of an embodiment of an imaging apparatus according to the present invention. The imaging apparatus 10 includes an imaging unit 11 that performs imaging of the imaging target 1, a processing unit 12 that performs processing of a video signal after imaging, and control that controls overall timing and parameters of the imaging unit 11 and the processing unit 12. Part 13.

  The imaging unit 11 includes an iris 111, a lens 112, a high-definition image sensor 113, and an analog / digital conversion unit 114. The iris 111, the lens 112, the high-definition image sensor 113, and the analog / digital conversion unit 114 are controlled by the control unit 13. The analog / digital conversion unit 114 converts the analog video signal read from the high-definition image sensor 113 into a digital video signal.

  Here, the high-definition image sensor 113 has four readout areas divided into four in the horizontal direction as shown by 21, 22, 23, and 24 in FIG. 2, and a video signal is read out for each readout area. It is an image sensor. Each of the readout areas 21, 22, 23, and 24 is scanned by the control unit 13 from the upper left to the lower right as indicated by 31 in FIG.

  On the other hand, as shown in FIG. 1, the processing unit 12 is a signal that performs basic video signal processing such as gamma processing and format conversion of a video signal, and a division boundary noise correction unit 121 that constitutes a main part of the present embodiment. And a processing unit 122. The division boundary noise correction unit 121 and the signal processing unit 122 are controlled by the control unit 13.

  FIG. 4 shows a block diagram of an embodiment of the division boundary noise correction unit 121. 4 is in charge of signal processing for one readout area in FIG. Therefore, the imaging apparatus 10 of the present embodiment has four systems of the configuration shown in FIG.

  In FIG. 4, an interpolation signal calculation circuit 41 creates an interpolation signal used for noise correction from an input video signal. This interpolation signal calculation circuit 41 assumes that the pixel to be interpolated in the input video signal is noise, and generates an interpolation signal by performing linear interpolation using the values of pixels before and after the pixel to be interpolated. To do.

  The memory 42 holds the input video signal for a time during which detection processing and interpolation processing are performed. The readout area front noise detection circuit 44 is configured to calculate the difference between the level of the target pixel and the level average value of each pixel adjacent to both sides before and after the target pixel with respect to the video signal in the range of several pixels in the front of the readout area. To detect high frequency components. The higher the high frequency component, the higher the possibility of noise.

  The value detected here is clipped at a constant level by the clip circuit 47, normalized by the normalization circuit 48 to a range of “0” to “1”, and stored in the memory 4E. The reason for clipping is to prevent false detection by an extremely high level signal and to prevent overcorrection when the amount of noise confirmed at the time of measurement is exceeded. The high frequency component detection range, the maximum and minimum levels to be clipped, and the range setting of levels corresponding to “0” to “1” at the time of normalization are determined by the read area front noise detection control unit 43. The Further, the read area front noise detection control unit 43 is set by the control unit 13 according to the noise characteristics investigated in advance.

  The readout area rear high luminance detection circuit 45 calculates an average value for the video signal in the range of several pixels in the rear part of the readout area of the immediately preceding line. The value calculated here is clipped at a certain level by the clipping circuit 4B for the above-described reason, and is normalized in the range of “0” to “1” by the normalizing circuit 4C and then stored in the memory 4D. The larger this value is, the higher the possibility that noise will appear at the front of the read area of the current line. The detection range, the maximum and minimum levels to be clipped, and the range setting of levels corresponding to “0” to “1” at the time of normalization are determined by the read area rear high luminance detection control unit 46. Further, the read area rear high luminance detection control unit 46 is set by the control unit 13 according to the noise characteristics investigated in advance.

  The readout area front noise detection signal normalized by the normalization circuit 48 and the readout area rear high luminance detection signal normalized by the normalization circuit 4C are multiplied by the multiplier 4F. This multiplied value is input to the selector 4A as a correction coefficient α. The selector 4A mixes the interpolation signal created by the interpolation signal calculation circuit 41 and the input video signal stored in the memory 42 according to the following equation at a ratio corresponding to the correction coefficient α and the correction coefficient (1-α). Thus, a corrected signal with a reduced noise component is generated.

Corrected signal = α × {interpolation signal} + (1−α) × {input video signal} (1)
The corrected signal output from the division boundary noise correction unit 121 having such a configuration is subjected to signal processing such as gamma correction and format conversion by the signal processing unit 122 in FIG. Is done.

  FIG. 5 shows an example of a video signal read from one of the divided read areas of the high-definition image sensor 113 in the image pickup apparatus 10. In the figure, the horizontal axis represents the horizontal position of the readout area, and the vertical axis represents the luminance. A solid line represents a signal in a state where noise is actually generated, and a broken line represents a signal on the immediately preceding line.

  The readout area front noise detection circuit 44 shown in FIG. 4 performs the noise detection and correction range (several pixels from the readout start horizontal position) 51 on the front several pixels of the readout area shown in FIG. The difference between the level (pixel value) of the pixel and the average value level (average pixel value) of both adjacent pixels is taken to extract a high frequency component. The higher the high frequency component, the higher the possibility of noise, but there is also a possibility that the picture is the actual image to be captured.

  Therefore, the readout region rear high luminance detection circuit 45 in FIG. 4 calculates an average value for the number of pixels in the rear portion of the immediately preceding line represented by a broken line in the high luminance detection range 52 shown in FIG. The higher this average value, the higher the possibility that noise is appearing in the noise detection range of the current line. Therefore, both detection results are combined with the noise detection and correction range 51 detection results for the front pixels of the readout area. The higher the value, the higher the value of the correction coefficient α output from the multiplier 4F.

  As a result, in the selector 4A, as can be seen from the expression (1), a relatively large amount of interpolation value components (interpolation signals) from the pixels adjacent to the processing target pixel (input video signal) are mixed. become. As a result, a high-intensity signal is input to the last read pixel at the boundary of the divided region of a certain line and the range of the last several pixels (the rear pixel), so that a range of several pixels from the first pixel of the next line ( When noise due to power supply voltage fluctuation due to a sudden change in signal or signal leakage due to signal wiring routing appears in the front pixel), noise detection and correction of the front pixel as shown in FIG. The noise component of the video signal in the range 61 is suppressed, and the image quality is improved. Further, in the present embodiment, since detection is performed at two locations, it is possible to prevent erroneous detection and correction for a normal video signal.

  The detection ranges 51 and 52 and the correction coefficient calculation level upper and lower limit values need to be measured in advance and set to optimum values in the control unit 13.

  Next, the operation of the present embodiment will be described with reference to the flowchart showing the processing procedure of FIG. When the user starts the operation (step S1), the control unit 13 in FIG. 1 sets the imaging unit 11 and the processing unit 12 in an optimum state depending on the imaging target 1 and imaging conditions (step S2). In addition, in the division boundary noise correction unit 121 of the processing unit 12, a noise detection position, a detection maximum / minimum level, a high luminance detection range, and a reading region front noise detection control unit 43 and a reading region rear high luminance detection control unit 46 are provided. The detection maximum / minimum level is set, and then imaging is started.

  First, the video signal of the line at the head of the frame is read (step S3). Subsequently, in the readout area rear high brightness detection control unit 46, the readout signal rear high brightness detection circuit 45 detects and averages the video signal levels of the rear part line pixels in the set range (step S4). Based on the maximum value / minimum value set by the area rear high luminance detection control unit 46, the normalization circuit 4C normalizes the range from “0” to “1” and then stores it in the memory 4D (step S5). ).

  Next, the video signal of the next line is read (step S6). Subsequently, the readout region front noise detection circuit 44 extracts a high frequency component from the video signal level of several pixels in the front part of the line within the range set by the readout region front noise detection control unit 43 (step S7), and further reads out. Based on the maximum value / minimum value set by the area front noise detection control unit 43, the normalization circuit 48 normalizes the range to “0” to “1” and then stores it in the memory 4E (step S8). ).

  Next, an interpolation value (interpolation signal) for noise correction is created by the interpolation signal calculation circuit 41 (step S9). The interpolation value (interpolation signal) is the average value of the video signal level of the pixels adjacent to the target pixel of the input video signal. Subsequently, the value calculated in step S5 and stored in the memory 4D is multiplied by the value calculated in step S8 and stored in the memory 4E by the multiplier 4F. The ratio between the input video signal and the value calculated in step S9 is changed and added by the selector 4a according to the equation (1), thereby outputting a corrected signal which is a video signal in which the noise component of the input video signal is corrected ( Step S10).

  Next, the post-correction signal output in step S10 is subjected to subsequent signal processing such as gamma correction and format conversion to obtain the final output of the imaging apparatus (step S11). Subsequently, it is determined whether or not the above process has reached the final line of the frame (step S12). If not, the process proceeds to the next line of the same frame, and the processes of steps S4 to S11 are repeated. If the final line has been reached, the process proceeds to the next frame (step S13), returns to step S3, and repeats the processes of steps S4 to S12 from the top line.

  Although not shown, when the user stops the operation, the process ends at that point and waits for the start of the next operation.

  Note that the present invention is not limited to the above-described embodiment. For example, the correction coefficient α supplied to the selector 4A may be represented by the following equation.

α = α1 × α2 (2)
However, in the above formula, α1 is calculated by the rear area high luminance detection circuit 45 by (rear pixel signal value−rear pixel noise appearance start signal value) / (maximum signal value−rear pixel noise appearance start signal value). It is a signal represented. Here, “the noise appearance start signal value of the rear pixel” is the value of the pixel that appears first among the pixels of the rear pixel and whose value is larger than a predetermined threshold value. The “maximum signal value” is a maximum value allowed as a pixel signal value. Α2 is a signal represented by (signal value of the target pixel of the front pixel) − (average value of signal values of pixels adjacent to the target pixel of the front pixel). However, α1 = 0 when α1 <0, and α2 = 0 when α2 <0.

  Further, in the above-described embodiment, the image pickup device in which the readout region is divided into a plurality of parts in the horizontal direction has been described. However, the present invention is not limited to this, and the present invention is similarly applied to an image pickup device divided in the vertical direction and in both the horizontal and vertical directions. The image quality can be improved by this process.

It is a block diagram of one embodiment of an imaging device of the present invention. It is a figure which shows an example of the imaging area of a high-definition image sensor in the imaging device of this invention. It is a figure which shows an example of the scanning method of the division | segmentation imaging area of a high-definition image sensor in the imaging device of this invention. FIG. 2 is a block diagram of an embodiment of a division boundary noise correction unit in the imaging apparatus of FIG. 1. It is a figure explaining the position and level of a noise detection and correction | amendment range and high-intensity detection range in one Embodiment of the imaging device of this invention. It is a figure explaining the position and level of the noise detection after noise correction and correction range in one embodiment of the imaging device of the present invention. It is a flowchart which shows the process sequence of one Embodiment of the imaging device of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image pick-up device 11 Image pick-up part 12 Signal processing part 13 Control part 21-24 Divided read area of high-definition image sensor 41 Interpolation signal calculation circuit 42, 4D, 4E Memory 43 Read-out area front noise detection control part 44 Read-out area front noise Detection circuit 45 Read area rear high brightness detection circuit 46 Read area rear high brightness detection control section 47, 4B clip circuit 48, 4C normalization circuit 4A selector 51 Before noise correction, noise detection and correction range 52 High brightness detection range 61 Noise correction Later, noise detection and correction range 113 High-definition image sensor 114 Analog / digital conversion unit 121 Division boundary noise correction unit 122 Signal processing unit

Claims (2)

An image pickup apparatus having a plurality of read areas obtained by dividing an image pickup area of an image sensor into a plurality, and reading video signals in parallel from each of the plurality of read areas,
Interpolation signal calculation means for calculating an interpolation signal used for noise correction by linear interpolation from the video signal output from the readout region;
In the first pixel range consisting of the first pixel of the correction target line of the video signal output from the readout area and a predetermined number of pixels immediately after that, the level of the target pixel and the target pixel A high-frequency component detection means for detecting a high-frequency component by taking a difference from the level average value of each pixel on both sides before and after
From the last read pixel of the boundary of the read area in the line immediately before one line detected by the high frequency component detecting means of the video signal output from the read area, and a predetermined number of pixels immediately before that Average value calculating means for calculating the average value of the video signal in the second pixel range,
Correction coefficient calculating means for calculating a correction coefficient by multiplying the high frequency component detected by the high frequency component detecting means and the average value calculated by the average value calculating means;
The relative addition ratio of the interpolation signal to the video signal output from the readout area is changed according to the value of the correction coefficient, and the corrected signal is generated by adding the video signal and the interpolation signal. An image pickup apparatus, comprising: a circuit unit comprising: an adding unit configured to perform a video signal output from each of the plurality of readout regions. Of each pixel in the second pixel range, the pixel signal value is a, the value of a pixel whose value is larger than a predetermined threshold value, and noise first appears in the second pixel range, b, When the maximum value as the pixel signal value is c, a coefficient calculating means for calculating a coefficient represented by (a−b) / (c−b) is provided instead of the average value calculating means,
The correction coefficient calculation means is means for calculating the correction coefficient by multiplying the coefficient calculated by the coefficient calculation means and the high frequency component detected by the high frequency component detection means. Item 2. The imaging device according to Item 1.

Images