JP6407686B2 - Noise reduction circuit - Google Patents

Description

  Embodiments described herein relate generally to a noise reduction circuit.

  In general, when a subject becomes dark, a video camera or the like can be photographed in a dark place by automatically adjusting a gain of an image by an AGC circuit built in the camera.

  By the way, when the gain of the AGC circuit is increased, the image becomes more noisy and the image becomes unsightly.

  For this reason, video cameras are equipped with noise reduction devices of various techniques to reduce noise.

  For example, a frame circulation type noise reduction device or the like attempts to reduce noise by extracting a noise component from a video difference between previous and next frames and subtracting the extracted noise component from the original video.

JP 2000-196916 A

  However, in the case of a conventional frame-circulating noise reduction device, if the noise reduction effect on moving images is increased, the difference between the previous and next video frames is used, resulting in a larger afterimage due to the motion of the video, resulting in lower video quality. There is a problem of doing.

  The problem to be solved by the present invention is to provide a noise reduction circuit capable of improving video quality by suppressing the occurrence of afterimages while reducing the noise of the video.

  The noise reduction circuit of the embodiment includes a delay unit, a first subtracter, a multiplier, a second subtractor, a limit circuit, and a control unit. The delay unit holds and delays the image one frame before the current image in units of frames. The first subtracter takes a difference between the current video frame and the previous video frame output from the delay unit, and generates a residual signal. The multiplier adjusts the level of the residual signal generated by the first subtracter according to a given feedback coefficient. The second subtracter outputs an image obtained by subtracting the residual signal whose level is adjusted by the multiplier from the current image. The limit circuit outputs a residual signal whose signal level exceeds a predetermined limit value by limiting it to the limit value. When the level of the residual signal input to the limit circuit exceeds a predetermined limit value and then falls below the limit value, the control unit rapidly decreases the feedback coefficient and gives it to the multiplier.

It is a figure which shows the structure of the video camera of embodiment. It is a figure which shows the structure of the noise reduction circuit of the video camera of FIG. It is a figure which shows an example of a table. It is a figure which shows the input / output characteristic of a limiter. It is a flowchart which shows operation | movement of a noise reduction circuit. It is a figure which shows an example of the output characteristic of KNR with progress of time. It is a figure which shows the afterimage reduction effect (comparative example) with respect to an input signal. It is a figure which shows the afterimage reduction effect (embodiment) with respect to an input signal.

Hereinafter, embodiments will be described in detail with reference to the drawings.
(Embodiment)
FIG. 1 is a diagram showing a configuration of a video camera according to one embodiment.
As shown in FIG. 1, the video camera according to the embodiment includes an imaging device 1, a sample and hold circuit (hereinafter, S / H circuit) 2, an A / D converter 3, an AGC circuit 5, a signal processing circuit 6, and an output format conversion. A circuit 7 is provided.

  The image sensor 1 captures an object, converts it into a video signal (electrical signal) (photoelectric conversion), and outputs it to the S / H circuit 2.

  The S / H circuit 2 samples and holds the electric signal for each pixel generated by photoelectric conversion in the image sensor 1 using a clamp pulse, and extracts a signal portion.

  The A / D converter 3 converts the signal portion sampled and held by the S / H circuit 2 into a digital video signal.

  The noise reduction circuit 4 generates a video signal from which noise has been removed or reduced from the digital video signal converted by the A / D converter 3 and outputs the video signal to the AGC circuit 5.

  The AGC circuit 5 amplifies the digital signal input from the noise reduction circuit 4 with a predetermined gain and outputs the amplified signal to the signal processing circuit 6.

  The signal processing circuit 6 performs signal processing such as gamma correction and detail processing on the digital signal amplified by the AGC circuit 5 with a predetermined gain, and outputs it to the output format conversion circuit 7 as a video signal.

  The output format conversion circuit 7 converts the video signal input from the signal processing circuit 6 into a standardized signal such as USB, DVI, or HDMI and outputs the signal to an external output terminal. (HDMI is a registered trademark.)

  As shown in FIG. 2, the noise reduction circuit 4 includes a subtracter 11 as a first subtracter, a limiter 12 as a limit circuit, a counter 13, a frame memory 14 for a counter, a control unit 15, a multiplier 16, a second An adder 17 as a subtracter, a frame memory 18 as a delay unit, a 1/64 multiplier 19 and the like are provided.

  The subtracter 11 subtracts (subtracts) the current video signal and the video signal one frame before (that is, the video signal delayed by one frame by the frame memory 18), and obtains a residual signal that is a noise component included in the video signal. Generate and output to the limiter 12. That is, the subtractor 11 generates a residual signal by taking the difference between the frame of the current video and the frame of the previous video output delayed by the frame memory 18.

  As shown in FIG. 3, the limiter 12 has input / output characteristics of an input signal versus an output signal.

  That is, the limiter 12 outputs a residual signal with a signal level exceeding a predetermined limit value max, limited to the limit value, and outputs the residual signal as it is when the signal level is equal to or lower than the limit value max.

  More specifically, the limiter 12 outputs the residual signal, which is an input signal from the subtractor 11, by limiting the residual signal to a predetermined level (limit value max) when the level is higher than a certain level, that is, a moving image. By limiting the feedback amount (in this case, the output of the subtractor 11 becomes large), it is possible to prevent an afterimage that is a harmful effect of noise reduction from increasing.

  When a completely stationary subject is imaged, only the noise component appears at the output of the subtractor 11, but when a moving object is imaged, the video signal of the previous frame and the current video signal are captured. Therefore, an afterimage is generated in order to add a signal including the video difference to the current video signal.

  The counter 13 decrements a predetermined numerical value (for example, a numerical value such as “15”) set in advance every predetermined time (every 1/60 sec) under the control of the control unit 15.

  In the counter frame memory 14, a decrement value counted by the counter 13 is set for each pixel of one frame.

  As shown in FIG. 4, in the table 10, the value of the coefficient KNR for the numerical value of the counter 13 (timer limit time) is set.

  The control unit 15 controls the counter 13 to start counting, reads the value of the coefficient KNR corresponding to the count value from the table 10, multiplies it by 1/64, and gives it to the multiplier 16 as the feedback system number K. In addition to 1/64, the value to be multiplied may be 1/128 or 1/256, for example. The resolution increases as the denominator of the value to be multiplied increases.

  When the level of the residual signal input to the limiter 12 drops below the limit value after the control unit 15 falls below the limit value, the control unit 15 multiplies the feedback system number K multiplied by 1/64 by rapidly decreasing the coefficient KNR. To the vessel 16. That is, the control unit 15 gives the feedback system number K to the multiplier 16 with a control characteristic such that the value of the coefficient KNR is gradually reduced and then gradually returned to the original value as time passes. The control characteristic (rising curve) is determined based on the relationship between the noise characteristic, the afterimage level, and the time priority.

The 1/64 multiplier 19 outputs the value of the feedback system number K obtained by multiplying the coefficient KNR input from the control unit 15 by 1/64 to the multiplier 16.
The multiplier 16 multiplies the residual signal through the limiter 12 by the feedback coefficient K given from the control unit 15 through the 1/64 multiplier 19 and outputs the result to the adder 17. That is, the multiplier 16 adjusts the level of the residual signal generated by the subtracter 11 according to the given feedback coefficient K. The feedback coefficient K is a value between 0 and 1 (0 ≦ K <1).

  As an example of the coefficient KNR that is a source of the feedback coefficient K, “53” is set as a predetermined limit value of the limiter 12, and “5” is set as a lower limit value when the limit is rapidly decreased.

  The adder 17 outputs a video signal obtained by subtracting the residual signal from the input video signal. That is, the adder 17 outputs a video obtained by subtracting the residual signal whose level has been adjusted by the multiplier 16 from the current video.

  The frame memory 18 holds the video signal input one frame before and outputs it to the subtracter 11 at the timing when the video signal of the next frame is input. That is, the frame memory 18 holds and delays the image one frame before the current image in units of frames.

  The noise reduction circuit 4 is a circuit that reduces the noise component of the video signal by utilizing the fact that the video signal has a correlation between frames and the noise component does not have a correlation between frames.

Next, processing in the video camera and operations of each unit will be described.
(Signal flow)
The video signal captured by the image sensor 1 is converted into an electrical signal by the image sensor 1 and input to the AGC circuit 5 via the S / H circuit 2. In the S / H circuit 2, a signal portion is extracted by sampling and holding from an electric signal for each pixel generated by photoelectric conversion in the image sensor 1 using a clamp pulse.

  The signal component extracted by the S / H circuit 2 is amplified by the AGC circuit 5 with a predetermined gain, then converted into a digital signal by the A / D converter 3, and input to the signal processing circuit 6, and the signal processing circuit 6 is subjected to signal processing such as gamma and detail processing, and is output as an image signal from an external terminal.

(Operation of noise reduction circuit 4)
The subtractor 11 subtracts the input current video signal and the video signal delayed by one frame by the frame memory 18 (one frame previous video signal) to generate a difference signal, that is, a residual signal, The generated residual signal is input to the multiplier 16 through the limiter 12.

  The multiplier 16 multiplies the input residual signal by a feedback coefficient K (0 ≦ K <1) and outputs the result to the adder 17. Note that the effect of noise reduction increases as the feedback coefficient K approaches 1, but the afterimage in a moving image increases. Therefore, the feedback coefficient K is set in accordance with the limit value of the limiter 12 and the performance and use conditions of the camera. It shall be.

  The adder 17 subtracts the residual signal input from the multiplier 16 from the current video signal, thereby outputting a video signal with reduced noise components.

  In this manner, the noise reduction circuit 4 of the video camera reduces the noise component of the video signal, and prevents the noise from being increased due to the gain increase in the next AGC circuit 5.

(Operation of control unit 15)
As illustrated in FIG. 5, when the residual signal level exceeds the limit value as a result of comparing the residual signal level and the limit value, the control unit 15 refers to the table 10 (step S101: Yes). The value A is set to the coefficient KNR and the value T is set to the count value TMR of the counter 13 (step S102), and the set coefficient KNR is output to the multiplier 16 through the 1/64 multiplier 19 (step S103). The process returns to the comparison process in step S101.

  Thereafter, when the limiter value exceeds and the residual signal level becomes equal to or lower than the limit value at a certain timing (No in step S101), the control unit 15 determines whether or not the count value TMR is 0 (step S101). S104) If the count value TMR is not 0 (No in step S104), the count value TMR is decremented (step S105), and the coefficient KNR of the table 10 corresponding to the count value TMR is multiplied through the 1/64 multiplier 19. 16 (step S106).

  For example, in the control characteristic diagram shown in FIG. 6, the level of the residual signal initially exceeds the limit value, and at the timings P1 to P4, the level of the residual signal immediately exceeds the limit value and immediately falls below the limit value. And

  In this case, at the timing P1 at time T1, the coefficient KNR is rapidly reduced from the max value “53” to the value A “5” (see the table 10 in FIG. 4), and this value “5” is multiplied by 1/64. (5 / 64≈0.078) is output to the multiplier 16. In addition, the count value TMR is set to a value T such as “15” (see the table 10 in FIG. 4).

  Thereafter, when the count value TMR of the counter 13 is decremented every 1/60 sec, the coefficient KNR of the table 10 gradually increases, and when the count value TMR of the counter 13 reaches 0, the coefficient KNR becomes the original value. Return to the max value “53”.

  After that, when the limiter value continues to exceed and the level of the residual signal becomes equal to or lower than the limit value at timing P2 at time T2, the coefficient KNR is rapidly reduced from the max value “53” to the value A “5”. A value obtained by multiplying / 64 (5 / 64≈0.078) is output to the multiplier 16.

  Note that the maximum value of the coefficient KNR, the suddenly decreasing value A, and the count value TMR are based on the type of video signal (moving image or still image, whether moving image is fast moving or slow moving), the amount of noise, and the like. The table 10 is set in consideration.

  After the coefficient KNR is rapidly reduced to the value A “5”, the level of the residual signal exceeds the limit value at the timing P3 at time T3 before the count value TMR becomes 0 while the coefficient KNR gradually increases. As soon as it becomes below the limit value, the coefficient KNR on the way up at that point is rapidly reduced to “5”.

  Then, the state where the residual signal level is below the limit value continues (period S from time T3 to time T4), and when the residual signal level exceeds the limit value and falls below the limit value at timing P4 at time T4, the counter A count of 13 is started.

  Then, at the timing P5 when the count value TMR of the counter 13 reaches 0, the coefficient KNR returns to the original max value “53”.

  Here, the difference in the afterimage reduction effect between the example in which the coefficient KNR is not adjusted (comparative example) and the present embodiment will be described. FIG. 7 shows the input / output characteristics of the video signal (afterimage component) when the coefficient KNR is not controlled (adjusted), and FIG. 8 shows the input / output characteristics of this embodiment.

  In the comparative example of FIG. 7, it can be seen that the time for which the output signal is delayed with respect to the input signal is long, there are many afterimage component regions S1 of the difference between the input signal and the output signal, and there are many afterimages, making it difficult to view as an image. On the other hand, in the present embodiment of FIG. 8, the time for which the output signal is delayed with respect to the input signal is short, the afterimage component region S2 of the difference between the input signal and the output signal is narrow, and the afterimage is difficult to see and the video is easy to see. You can see that

  As described above, according to the video camera of this embodiment, the coefficient KNR (or the feedback coefficient K) is obtained when the residual signal that is the difference between the preceding and succeeding video signals exceeds the limit value and then returns to the limit value range again. By temporarily controlling to a small value, it is possible to reduce the afterimage while reducing the noise component of the image. In other words, video quality can be improved by suppressing the occurrence of afterimages while reducing video noise.

  By keeping the setting of the limiter 12 fixed, that is, without adjusting it, the feedback coefficient K of the multiplier 16 can be varied to improve the follow-up performance with respect to changes in the image.

  Although the embodiment of the present invention has been described, this embodiment is presented as an example and is not intended to limit the scope of the invention. The novel embodiment can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  In the above embodiment, the video signal is held in units of frames, but may be in units of fields. Further, in the above embodiment, the counter frame memory 14 and the video delay frame memory 18 are provided separately, but may be one.

  In the above embodiment, only one relationship between the time and the coefficient is set in the table 10. However, the table 10 stores a plurality of settings for each shooting mode and changes the feedback system number K according to the shooting mode. May be changed.

  Furthermore, in the above-described embodiment, the example in which the control unit 15 and the 1/64 multiplier 19 are separately provided has been described. However, the control unit 15 includes the function of the 1/64 multiplier 19 and feedback from the control unit 15. The coefficient K may be directly input to the multiplier 16.

  Each component shown in the above embodiment may be realized by a program installed in a storage such as a hard disk device of a computer, and the program is stored in a computer-readable electronic medium: program. The computer may realize the functions of the present invention by causing the computer to read from the electronic medium. Examples of the electronic medium include a recording medium such as a CD-ROM, flash memory, and removable media. Further, the configuration may be realized by distributing and storing components in different computers connected via a network, and communicating between computers in which the components are functioning.

  DESCRIPTION OF SYMBOLS 1 ... Imaging device, 2 ... Sample hold circuit (S / H circuit), 3 ... A / D converter, 4 ... Noise reduction circuit, 5 ... AGC circuit, 6 ... Signal processing circuit, 7 ... Output format conversion circuit, DESCRIPTION OF SYMBOLS 10 ... Table, 11 ... Subtractor, 12 ... Limiter, 13 ... Counter, 14 ... Frame memory (for counter), 15 ... Control part, 16 ... Multiplier, 17 ... Adder, 18 ... Frame memory (for delay), 19: 1/64 multiplier.

Claims (3)

A delay unit that holds and delays the image one frame before the current image in units of frames;
A first subtracter that generates a residual signal by taking a difference between the frame of the current video and the frame of the previous video output from the delay unit;
A multiplier for adjusting a level of the residual signal generated by the first subtractor according to a given feedback coefficient;
A second subtracter for outputting a video obtained by subtracting the residual signal whose level is adjusted by the multiplier from the current video;
A limit circuit that outputs the residual signal with a signal level exceeding a predetermined limit value, limited to the limit value;
A noise reduction unit comprising a control unit that rapidly reduces the feedback coefficient and supplies the multiplier to the multiplier when the level of the residual signal input to the limit circuit falls below a limit value after exceeding a predetermined limit value. circuit. The controller is
The noise reduction circuit according to claim 1, wherein the feedback coefficient is given to the multiplier with a control characteristic such that the feedback coefficient is gradually decreased and then gradually returned to the original value as time passes.   The noise reduction circuit according to claim 2, wherein the control characteristic is determined based on a relationship among a noise characteristic, an afterimage level, and a time priority.

Images