JP5403784B2 - Video processing apparatus and noise reduction method - Google Patents

Description

  The present invention relates to a video processing apparatus including means for reducing noise superimposed on an acquired video signal, and a noise reduction method using the video processing apparatus.

  In recent years, many video processing apparatuses (TV receivers, etc.) having a noise reduction (hereinafter referred to as “NR”) function as means for reducing noise superimposed on a video signal in a transmission path or the like have been seen. In NR, a video signal of a rear frame is output as “a video signal obtained by mixing video signals output at the same pixel position in two consecutive frames before and after at a predetermined rate”, and superimposed on the video signal of the rear frame. Technologies such as 3DNR that reduce the generated noise have been developed, and by using these functions, it is possible to output an image with less noise.

  However, when the 3DNR is operated in a scene with motion, an afterimage is generated in the output video, and a bad effect such as a blur is generated. Specifically, in the video scene in which the character's face moves as shown in FIG. 1A, if 3DNR is operated, noise can be reduced as shown in FIG. The face of the character appears blurry when the image of the back frame is output (afterimage). The problem of the afterimage can be solved by weakening the effect of 3DNR. However, as a matter of course, there is a problem that noise cannot be sufficiently reduced if the 3DNR effect is weakened.

  That is, if the 3DNR effect is strengthened, noise can be sufficiently reduced, but an afterimage generated in a moving scene becomes conspicuous. On the other hand, if the 3DNR effect is weakened, an afterimage generated in a moving scene becomes inconspicuous, but noise cannot be sufficiently reduced.

  Therefore, as means for solving the above problem, means for dynamically switching the effect of 3DNR according to the video scene is known. Specifically, the difference between the acquired video signals of two consecutive frames before and after is calculated, and if the calculated difference is large, the 3DNR is not operated, and if the difference is small, the 3DNR is operated with a certain effect. .

  That is, when the difference between the video signals is large, it is determined that the scene has a motion, and the 3DNR is not operated, thereby preventing the occurrence of an afterimage. On the other hand, when the difference between the video signals is small, it is determined that the scene does not move much, and the 3DNR is operated with a certain effect to reduce noise. In the prior art, as described above, even when the difference is small, 3DNR is operated only with a certain effect. This is because, since it is not possible to clearly determine whether the difference is due to noise or switching of the video signal, a countermeasure is taken so as not to generate an afterimage. An invention provided with such means is described in Patent Document 1, for example.

  Certainly, the prior art effectively solves the afterimage problem by dynamically switching the effect of 3DNR. However, since the 3DNR is not operated with a sufficient effect, there remains a problem that the noise problem cannot be sufficiently solved (see FIG. 1C).

  Here, when the source recorded on the movie film is used in television broadcasting or the like, a conversion method such as “2: 2 pulldown” or “2: 3 pulldown” is used.

“2: 3 pull-down” is a method of converting a movie film source recorded at 24 frames per second into a 60 Hz video signal as a source for television broadcasting. That is, the video of 24 frames is converted into a video signal of 60 frames. Specifically, as shown in FIG. 2, “one video (N _n ) of one frame of movie film is continuous 2 Use as video for frame → use video of next frame (N _{n + 1} ) as video for 3 consecutive frames → use video of next frame (N _{n + 2} ) as video for 2 consecutive frames → further The frame image (N _{n + 3} ) is used as an image for three consecutive frames →... The conversion method is used in Japan and the like.

“2: 2 pulldown” is a method of converting a movie film source recorded at 24 frames per second into a 50 Hz video signal as a source for television broadcasting. That is, the video of 24 frames is converted into a video signal of 50 frames. Specifically, as shown in FIG. 3, “video (N _n ) of one frame of movie film is continuous 2 Use as video for frame → Use video for next frame (N _{n + 1} ) as video for 2 consecutive frames → Use video for next frame (N _{n + 2} ) as video for 2 consecutive frames → Further next The frame image (N _{n + 3} ) is converted as a continuous two-frame image →... This conversion method is used in Europe and the like.

In this way, a video source (hereinafter referred to as “pull-down source”) converted by the “2: 2 pull-down” or “2: 3 pull-down” method always has two or three consecutive frames of the same video. Will appear. In such a case, the difference between the video signals detected between frames of the same video is 100% due to noise, and an afterimage does not occur even when 3DNR is operated with a strong effect using the two frames. Absent. If the 3DNR operation can be dynamically switched in consideration of this point, it is considered that a sufficient noise reduction effect can be realized without causing the afterimage problem and compared with the prior art. It is done.
JP 2006-197455 A

  Therefore, in the present invention, when a video source having undergone conversion such as “2: 2 pulldown” or “2: 3 pulldown” is acquired, the operation of 3DNR is appropriately controlled by an unconventional method, and thus an afterimage is obtained. It is an object of the present invention to provide a video processing apparatus configured to realize a sufficient noise reduction effect without causing the above problem.

  As means for solving the above problems, the following inventions are provided.

  When the acquired video source is a pull-down source converted such as “2: 2 pull-down” or “2: 3 pull-down”, and the determination result is a pull-down source, Provided is a video processing apparatus that appropriately controls the operation (effect) of 3DNR according to a repetition rule of the same video frame in a video source.

  Specifically, the following inventions are provided.

  In the first invention, a 3DNR unit that performs processing to reduce noise superimposed on the video signal, a determination unit that determines whether the acquired video source is a pull-down source, and a determination result that the determination result is a pull-down source. In some cases, there is provided a video processing apparatus including a noise reduction amount determination unit that determines a noise reduction amount in a 3DNR unit according to a repetition rule of the same video frame in a video source.

  The second invention is based on the first invention, and the noise reduction amount determination unit further includes a determination unit for determining a pull-down source type, and two consecutive identical videos that appear when the determination result is 2: 3 pull-down. Alternatively, first dynamic control means for determining a noise reduction amount between the same images of three consecutive identical images as a maximum threshold value, and a noise reduction amount as a value corresponding to the image change amount between the changed images straddling the same continuous video group. An image processing apparatus having the above is provided.

  The third invention is based on the first invention or the second invention, and the noise reduction amount determination unit appears when the determination means for determining the pull-down source type and the determination result is 2: 2 pull-down. A second dynamic control means for determining a noise reduction amount between the same images of two consecutive identical images as a maximum threshold value, and a noise reduction amount between the changed images straddling the continuous identical image group as a value corresponding to the image change amount; A video processing apparatus is provided.

  In the fourth invention, when the determination step is to determine whether the acquired video source is a pull-down source and the determination result is a determination result that the determination is a pull-down source, the repetition rule of the same video frame in the video source is followed. A video processing apparatus comprising: a noise reduction amount determining step for determining a noise reduction amount; and a 3DNR step for performing a process for reducing noise superimposed on the video signal based on the noise reduction amount determined in the noise reduction amount determination step. Provides a noise reduction method.

  When a video source converted such as “2: 2 pulldown” or “2: 3 pulldown” is acquired by the video processing apparatus and the noise reduction method of the video processing apparatus of the present invention, the afterimage problem does not occur. In addition, a sufficient noise reduction effect can be realized.

Examples of the present invention will be described below. In addition, this invention is not limited to these Examples at all, and can be implemented in various modes without departing from the gist thereof.
<Overview of embodiment>

  The video processing apparatus according to the present embodiment has a 3DNR function for reducing noise superimposed on the acquired video signal. Further, when a video source is acquired, the video source has a function of determining whether or not the video source is a pull-down source that has undergone conversion such as 2: 3 pull-down or 2: 2 pull-down.

  When it is determined that the acquired video source is a pull-down source, the effect of noise caused by 3DNR is dynamically changed according to the repetition rule of the same video frame determined by a conversion method such as 2: 3 pull-down or 2: 2 pull-down. It is characterized by switching. Specifically, 3DNR is operated with a sufficiently large effect between frames in which the same video frame is continuous, and 3DNR is not operated between frames in which different video frames are continuous, or operates with a certain effect. Let

  Here, the effect when the 3DNR is operated by acquiring the pull-down source by the video processing apparatus according to the present embodiment will be described with reference to FIGS. The figure shows a video scene in which the character's face moves, and the time direction is indicated by arrows.

  FIG. 1A shows a state in which video is output to the screen using the acquired video source as the original signal. In such a case, noise superimposed on the video signal becomes conspicuous.

  FIG. 1B shows a state in which the video is output to the screen after the 3DNR is uniformly operated on the acquired video source. In such a case, the noise superimposed on the video signal is sufficiently reduced and becomes inconspicuous, but the afterimage of the face of the moving character becomes conspicuous.

  FIG. 1C shows a state in which video is output to the screen after 3DNR is operated (prior art) while dynamically switching the effect on the acquired video source. In such a case, an afterimage of a moving character's face does not occur, but noise superimposed on the video signal cannot be sufficiently reduced and becomes noticeable.

FIG. 1D shows a state in which the video is output to the screen after the 3DNR is operated on the video source which is the acquired pull-down source by the video processing apparatus of the present embodiment. In such a case, the noise superimposed on the video signal can be sufficiently reduced without causing an afterimage of the face of the character in motion.
<Functional configuration of this embodiment>

  FIG. 4 is a diagram illustrating an example of functional blocks of the video processing apparatus according to the present embodiment. As shown in this figure, the “video processing apparatus” (0400) of this embodiment includes a “3DNR unit” (0401), a “determination unit” (0402), and a “noise reduction amount determination unit” (0403). . The “noise reduction amount determination unit” (0403) may include a “determination unit” and a “first dynamic control unit”. Furthermore, the “noise reduction amount determination unit” (0403) may include “second dynamic control means”.

  Note that the functional blocks of the apparatus described below can be realized as hardware, software, or both hardware and software. Specifically, if a computer is used, a CPU, a RAM, a bus, or a secondary storage device (a storage medium such as a hard disk, a non-volatile memory, a CD-ROM or a DVD-ROM, and a read drive for the medium) ), Hardware components such as printing devices, display devices, and other external peripheral devices, I / O ports for the external peripheral devices, driver programs for controlling these hardware, other application programs, and information input User interface.

  In addition, these hardware and software process the programs developed on the RAM with the CPU, and process, store, and output the data stored on the memory and hard disk, and the data input via the interface. Or used to control each hardware component. The present invention can be realized not only as an apparatus but also as a method. A part of the invention can be configured as software. Furthermore, a software product used for causing a computer to execute such software, and a storage medium in which the product is fixed to a storage medium are naturally included in the technical scope of the present invention (the same applies throughout this specification). Is).

  Hereinafter, the “3DNR unit” (0401), the “determination unit” (0402), the “noise reduction amount determination unit” (0403), and the “noise reduction amount determination unit” of the “video processing apparatus” (0400) of this embodiment ( The functional configuration of “determination means”, “first dynamic control means”, and “second dynamic control means” of 0403) will be described.

  The “video processing device” is a device that performs predetermined processing for outputting video to the screen with respect to the acquired video source, and corresponds to, for example, a television receiver. As a means for acquiring a video source, a TV broadcast wave is acquired via a tuner, a video source stored in another external device, or a video source read from a recording medium such as a DVD by another external device is used. A means for obtaining by wired or wireless communication is conceivable.

  The “3DNR unit” (0401) is configured to perform processing to reduce noise superimposed on the video signal. Specifically, it is configured to reduce noise superimposed on the video signal by using a technique of “3DNR (3D noise reduction)”. “3DNR” is a process of removing noise components using video signals of two consecutive frames. Specifically, “3DNR” is a process of “video signals output to the same pixel position of two consecutive frames”. This corresponds to a process of reducing the noise superimposed on the video signal of the subsequent frame by outputting the video of the subsequent frame with the “video signal mixed at a predetermined rate”.

  Here, the 3DNR unit (0401) of the present embodiment reduces the 3DNR noise reduction amount according to “difference information” that is the sum of the differences between the video signals output to the same pixel position in two consecutive frames. There may be provided means for performing processing to reduce noise while dynamically switching. The “noise reduction amount” is a concept for expressing the effect of noise reduction by 3DNR, and does not specifically indicate the amount of reduced noise in any unit. The expression method is an arbitrary design matter. For example, as a 3DNR, a video signal in a subsequent frame is a video signal obtained by mixing video signals output at the same pixel positions in two consecutive frames before and after at a predetermined rate. When the output process is performed, the “noise reduction amount” may be expressed by a ratio of mixing the video signal of the previous frame. Specifically, when the video signal of the previous frame is mixed by 20%, “the noise reduction amount is 20%”, and when the video signal of the previous frame is mixed by 50%, the “noise reduction amount is 50%”. % ”And“ Noise reduction amount is 0% ”when the video signal of the previous frame is not mixed at all (the“ noise reduction amount ”is the concept).

  Note that the process of dynamically switching the 3DNR noise reduction amount according to the difference information increases the noise reduction amount to some extent when the difference is small. The noise reduction amount is an arbitrary design item, but is preferably about 10 to 30% in view of problems such as the occurrence of afterimages. On the other hand, when the difference is large, the noise reduction amount is reduced. The noise reduction amount is also an arbitrary design item, but it is desirable to set it to about 0 to 10% in consideration of problems such as the occurrence of afterimages. Whether the difference is large or small is determined based on whether the difference is larger or smaller than a threshold (arbitrary design item) held in advance.

  The “determination unit” (0402) is configured to determine whether the acquired video source is a pull-down source. The “pull-down source” is a video source converted by the “2: 3 pull-down” or “2: 2 pull-down” method. Since “2: 3 pull-down” and “2: 2 pull-down” are as described above with reference to FIGS. 2 and 3, detailed description thereof is omitted here.

  Here, as a means for determining whether or not the acquired video source is a pull-down source, for example, “2: 3 pull-down” or “2: 2 pull-down” specific “same video frame repetition” is included in the acquired video source. It may be realized by monitoring whether the same video frame appears according to the “Rule”.

  As a specific means, for example, each time a video signal for one frame is acquired, it is determined whether or not it is the same as the video signal of the immediately preceding frame, and the result (“same” or “no”) is stored. I will do it. Then, it may be realized by determining whether the transition of the stored result does not correspond to a repetition rule specific to “2: 3 pulldown” or “2: 2 pulldown”.

  In other words, transition of results such as “different → same → different → same → same → different → same → different → same → same → different → ...” (different (once) → same (once) → different When (1 time) → same (2 times) ”is assumed to be one cycle and the cycle repeatedly appears), it is determined that the source is a pull-down source converted by the“ 2: 3 pull-down ”method. More specifically, the transition for one cycle of “different (once) → same (once) → different (once) → same (once)” is continuously performed a predetermined number of times (once, twice). (5 times, etc.), it may be determined that the source is a pull-down source converted by the “2: 3 pull-down” method.

  In addition, the transition of the result such as “different → same → different → same → different → same →…” (“different (1 time) → same (1 time)” is one cycle, and the cycle appears repeatedly. ) Appears, it is determined as a pull-down source converted by the “2: 2 pull-down” method. More specifically, the transition of one cycle “different (once) → same (once)” is repeated a predetermined number of times (once, twice, five times, etc.). If confirmed, it may be determined that the source is a pull-down source converted by the “2: 2 pull-down” method.

  Note that “determining whether the video signal of the acquired frame is the same as the video signal of the immediately preceding frame” is an arbitrary design matter, for example, calculating the average pixel value of the video signal of each frame, The process may be the same when the difference is smaller than the threshold stored in advance, and different when it is larger than the threshold. Alternatively, the difference information between the video signals of the two frames may be calculated, and the process may be determined to be the same when the value is smaller than the threshold stored in advance and different when the value is larger than the threshold. The threshold is an arbitrary design item.

  When the determination unit (0402) determines whether or not the acquired video source is a pull-down source, the determination unit (0402) notifies the “noise reduction amount determination unit” (0403) described below of the result. In addition, when the determination unit (0402) determines that the source is a pull-down source, the determination unit (0402) is converted by any type of method ("2: 3 pull-down" or "2: 2 pull-down") together with the result. Information identifying whether the source is a pull-down source may also be notified.

  Here, as a timing at which the determination unit (0402) determines whether or not it is the pull-down source, for example, the determination unit (0402) may always perform while acquiring the video source, or may perform it every predetermined time. Good. In the case of always performing, it is determined whether or not the video signal of all the frames of the acquired video source is the same as the video signal of the immediately preceding frame, and the transition of the determination result is “2: 3 pulldown” or “ Whether or not the transition is unique to “2: 3 pull-down” will always be monitored. The result is notified to the “noise reduction amount determination unit” (0403) described below at all times or whenever the determination of whether or not the source is a pull-down source and the determination of the pull-down source type are switched. May be.

  In addition, when it is performed every predetermined time (arbitrary design matter, “Example: 10 seconds, 30 seconds, 1 minute, etc.”), a predetermined number of continuous times acquired after that timing triggered by the passage of the predetermined time Using a video signal of a frame (arbitrary design item, “e.g., 8 frames, 10 frames, 20 frames, etc.”), it is determined whether or not it is the same as the video signal of the previous frame. Then, it is determined whether or not the acquired video source is a pull-down source by determining whether or not the transition of the determination result is a transition specific to “2: 3 pull-down” or “2: 2 pull-down”. Each time the above processing is performed or each time the determination of whether or not the source is a pull-down source and the determination of the pull-down source type are switched, the result is notified to a “noise reduction amount determination unit” (0403) described below. You may do it.

  If the determination result is a pull-down source, the “noise reduction amount determination unit” (0403) determines the noise reduction amount in the 3DNR unit according to the repetition rule of the same video frame in the video source. Configured to determine. The “repetition rule of the same video frame in the video source” is a predetermined rule in which the same video frame appears repeatedly. In the case of “2: 3 pull-down source”, as shown in FIG. In the case of “2: 2 pull-down source”, “2 → 2 → 2 →...” As shown in FIG. It is.

  That is, the noise reduction amount determination unit (0403) determines to dynamically switch the noise reduction amount in the 3DNR unit according to the repetition rule of the same video frame according to the type of pull-down source. In this case, when the determination result is a pull-down source, the noise reduction amount determination unit (0403) (a) means for determining which type the pull-down source is, B) It is necessary to include means for dynamically determining the noise reduction amount in the 3DNR unit according to the same video frame repetition rule according to the determined type. Therefore, in order to realize the above, the noise reduction amount determination unit (0403) may include “determination means” and “first dynamic control means”. Further, “second dynamic control means” may be included.

  The “determination means” is configured to determine the pull-down source type. Specifically, it is determined whether the acquired pull-down source corresponds to “2: 3 pull-down source” or “2: 2 pull-down source”. As a specific means of determination, for example, when a notification that “the acquired video signal is a pull-down source” is received from the determination unit (0402), along with information indicating that, the acquired video source Also, information indicating whether the source is a pull-down source converted by the method of the type is acquired. And you may make it determine by analyzing the information. The means for the determination unit (0402) to determine by which method the acquired video source has been converted is as described above.

  When the determination result is 2: 3 pull-down, the “first dynamic control unit” uses the same amount of noise reduction amount as the maximum threshold between the same video of two consecutive identical videos or three consecutive identical videos that appear. Between the change images straddling the image groups, the noise reduction amount is determined to be a value corresponding to the image change amount. “Two consecutive identical images” means that frames of the same image appear twice in succession, and in the case of the source after 2: 3 pulldown in FIG. 2, the images of the first and second frames, The 6th and 7th frame images are applicable. That is, “between the same images of two consecutive identical images” means between the frames of the same image, and in the case of FIG. 2, between the first frame and the second frame, and between the sixth frame and the seventh frame. It is between things.

  Similarly, in the case of the source after the 3: 3 pull-down in FIG. 2, “3 consecutive identical videos” corresponds to the 3rd, 4th, and 5th frames. In the case of FIG. 2, “between images” means between the third frame and the fourth frame, between the fourth frame and the fifth frame, and the like.

  The first dynamic control means determines a noise reduction amount to be a predetermined maximum threshold between two consecutive identical videos that appear or between identical videos between three consecutive identical videos. The “noise reduction amount between the same images” is a noise reduction amount of 3DNR performed by using the video signals of two frames constituting the same image, that is, the rear side constituting the same image. This is a noise reduction amount of 3DNR for reducing noise superimposed on the video signal of the frame. The maximum threshold value is an arbitrary design item, and can be set to 50%, 70%, 90%, or the like, for example. Even when such a high numerical value is set, the problem of afterimage does not occur.

  Next, “between the changed videos that straddle the same continuous video” means a frame between the borders of two or more consecutive same video groups that are the same video or three consecutive same videos. For later sources, it is between the second and third frames, between the fifth and sixth frames, and so on. That is, “between the changed video images that straddle the same continuous video group” refers to between different video frames. The first dynamic control means sets the noise reduction amount to a value corresponding to the image change amount between the change images straddling the same continuous video group. For example, a predetermined noise reduction amount based on the difference information is determined. The means is the same as described in the 3DNR section (0401). Note that the first dynamic control means determines the noise reduction amount to a sufficiently small threshold value (eg, “0%, 5%, etc.”) that is set in advance between the change videos that straddle the same video group. You may do it.

  Here, the effect when the noise reduction amount is dynamically changed by the first dynamic control means will be described with reference to FIG. FIG. 5 shows a case in which the 3DNR effect (noise reduction amount) is dynamically switched (upper side in the figure) by the conventional technique with respect to the 2: 3 pull-down source, and the 3DNR effect (noise) using the present invention. This is a side-by-side view of the case where the reduction amount is dynamically switched (lower side in the figure).

  As shown in the figure, in the conventional technique (upper side in the figure), 3DNR is operated with a certain amount of noise reduction (eg, 0 to 30%) according to the difference information between any frames. In such a case, the problem of afterimage does not occur, but sufficient noise reduction cannot be realized.

  On the other hand, in the present invention (lower side in the figure), a strong noise reduction amount (for example, between the first frame and the second frame, between the third frame and the fourth frame, etc.) 70%), 3DNR is operated and noise is sufficiently reduced. Needless to say, no matter how much the noise reduction amount is increased between frames of the same video, the problem of afterimage does not occur.

  Then, between different video frames (between the second and third frames, between the fifth and sixth frames, etc.), a certain amount of noise reduction according to the difference information (example: 0 to 30%) To operate 3DNR. As a result, the problem of afterimage is prevented.

  As described above, when the 3DNR effect (noise reduction amount) is dynamically switched using the present invention (lower side in the figure), the afterimage problem does not occur between any frames. . In addition, since 3DNR is operated with a sufficiently strong effect between frames of the same video, 3DNR is operated only with a certain amount of noise reduction (for example, 0 to 30%) according to difference information between any frames. An excellent noise reduction effect can be obtained as compared with the conventional technology that does not.

  Note that the numerical values of noise reduction (“0 to 30%” and “70%”) described in the above description and the numerical values in FIG. 5 (“0 to 30%” and “70%”) are merely examples. Yes, it is not limited to these.

  When the determination result is 2: 2 pull-down, the “second dynamic control unit” changes the amount of noise reduction between the same images of the two consecutive identical videos that appear across the consecutive identical video groups. Between images, the noise reduction amount is set to a value corresponding to the image change amount. Note that the concepts of “two consecutive identical videos” and “between identical videos of two consecutive identical videos” are the same as described above. Further, the maximum threshold value of the noise reduction amount is an arbitrary design item, and can be 50%, 70%, 90%, or the like, for example. Furthermore, “between change videos that straddle the same continuous video group” means a frame between the borders of two continuous same video groups that are the same video, and other concepts (such as “between change video”) are described above. It is the same.

  Even when the noise reduction amount is dynamically changed by the second dynamic control means, the same excellent effect as when the noise reduction amount is dynamically changed by the first dynamic control means (FIG. 5). Can be obtained.

  <Hardware configuration of this embodiment>

  FIG. 6 is a diagram illustrating an example of a configuration when the functional configuration of the video processing apparatus is realized as hardware. As shown in the figure, the video processing apparatus of the present embodiment includes a “3DNR unit”, “determination unit”, “noise reduction amount determination unit”, “determination means” of the “noise reduction amount determination unit”, “first motion”. "CPU" (0601), "main storage device" (0602), "program storage device" (0603), "secondary storage device" (0604) constituting "dynamic control means", "second dynamic control means", etc. ), “VRAM” (0605), “User I / F” (0606), “Tuner” (0607), “External Device I / F” (0608), “Display” (0609), “Speaker” (0610) , Etc., which are connected to each other by a data communication path such as “system bus” (0611), and perform transmission / reception and processing of information.

  Hereinafter, an example of means for realizing the function of the video processing apparatus according to the present embodiment will be described with reference to the hardware diagram of FIG.

  First, according to the broadcast wave acquisition program developed in the work area of the main storage device (0602), when a video source is acquired via the tuner (0607), it is converted into a YCbCr signal and stored in the data area of the main storage device (0602). Store. For example, when a video source is acquired via a digital tuner, the MPEG signal is converted into a YCbCr signal by decoding processing or the like. When a video source is acquired via an analog tuner, it is converted into a YCbCr signal by YC separation processing or A / D conversion.

  In addition, even when a video source is acquired from an external device via an external device I / F (0608) such as an HDMI terminal, a CVBS terminal, or a component terminal, it is converted into a YCbCr signal using the conventional technology.

  Thereafter, according to the NR program, NR processing is performed on the YCbCr signal for one frame (for example, 2DNR processing may be performed on the signal of the first frame, or no processing is performed). The signal after NR processing is stored in the data area of the main memory (0602). Then, in accordance with the video output program, predetermined processing such as IP conversion is performed on the signal after NR processing for one frame, and then the signal is written in the VRAM (0605) to output the video to the display (0609). .

  Also, NR processing is performed on the YCbCr signal of the next frame, and the signal after NR processing is stored in the data area of the main storage device (0602). The NR process uses the post-NR signal of the previous frame stored in the data area of the main storage device (0602), and the 3DNR process with the noise reduction amount corresponding to the difference information of the video signals of the two frames It may be. Thereafter, the signal after NR processing is compared with the signal after NR processing of the previous frame, it is determined whether or not the images of these two frames are the same, and the determination result is stored in the data area of the main memory (0602). To store. The determination may be, for example, a process of calculating an average pixel value of each of two frames and determining whether or not a difference between them is larger than a predetermined value held in advance. Further, after performing predetermined processing such as IP conversion on the NR processed signal for the frame in accordance with the video output program, the signal is written to the VRAM (0605) and the video is output to the display (0609). .

  When the above process is repeated and a predetermined number of determination results as to whether or not the images of two consecutive frames are the same are accumulated, the transition of the accumulated determination results is a transition peculiar to the 2: 3 pull-down source held in advance. Or, a comparison process is performed to determine whether the transition is unique to the 2: 2 pull-down source. As a result of the comparison, for example, when it coincides with the transition specific to the 2: 3 pull-down source, the YCbCr signal acquired thereafter is identical to the 2: 3 pull-down source according to the 2: 3 pull-down source NR program. 3DNR is performed while the noise reduction amount is dynamically switched according to a video frame repetition rule. The processing after NR processing is the same as described above.

Thereafter, each time a signal after NR processing is acquired, it is determined whether or not it is the same as the signal after NR processing of the previous frame, and the result is accumulated. Moreover, it continues to compare whether the transition of the accumulated determination result matches the transition unique to the 2: 3 pull-down source. As a result of the comparison, when the transition of the determination result does not coincide with the transition specific to the 2: 3 pull-down source, the NR processing of the previous frame is performed on the YCbCr signal obtained thereafter according to the NR program. Using the post signal, 3DNR processing or the like is performed with a noise reduction amount corresponding to the difference information between the video signals of the two frames. The processing after NR processing is the same as described above.
<Processing flow of this embodiment>

  An example of the processing flow of the video processing apparatus of this embodiment will be described with reference to the flowchart of FIG.

  First, when the video processing apparatus according to the present embodiment acquires a video source via a tuner or the like (S0701), it is determined whether the acquired video source is a pull-down source (2: 3 pull-down source or 2: 2 pull-down source). Judgment is made (S0702).

  If the result of the determination is that the video source is not a pull-down source (S0703), two consecutive frames are compared and a noise reduction amount is dynamically determined according to the difference information (S0705). Note that “when it is not a pull-down source” in step S0703 includes a case where it is not possible to determine whether or not it is a pull-down source (for example, when a predetermined number of frames of video signals are not acquired).

  On the other hand, when the video source is a pull-down source (S0703), the noise reduction amount is dynamically determined according to the same video frame repetition rule specific to the pull-down source (S0704). Specifically, the noise reduction amount is determined as the maximum threshold between the same images of two consecutive identical videos or three consecutive identical videos appearing in the pull-down source. In addition, between the change videos that straddle the same video group that is two consecutive identical videos or three consecutive identical videos, the image change amount is determined according to the image conversion amount between them. Thereafter, while the 3DNR is operated with the determined noise reduction amount (S0706), the image subjected to the noise reduction processing is output to the screen.

Until the power of the video processing apparatus is turned off (S0707), it is determined whether or not the acquired video source is a pull-down source (S0702), and means for determining the 3DNR noise reduction amount is activated according to the determination result. The 3DNR is operated while changing the image, and the image subjected to the noise reduction processing is output on the screen (S0703 to S0706).
<Effect of this embodiment>

  When a video source converted such as “2: 2 pull-down” or “2: 3 pull-down” is acquired by the video processing apparatus and the noise reduction method of the video processing apparatus according to the present embodiment, a problem of afterimage occurs. And a sufficient noise reduction effect can be realized.

Conceptual diagram 1 for explaining the effect of the noise reduction method of the present invention Conceptual diagram for explaining 2: 3 pull-down conversion Conceptual diagram for explaining 2: 2 pull-down conversion Example of functional block diagram of video processing apparatus of the present invention Conceptual diagram 2 for explaining the effect of the noise reduction method of the present invention The conceptual diagram which showed an example of the hardware constitutions of the video processing apparatus of this invention The flowchart figure which shows an example of the flow of a process of the video processing apparatus of this invention

Explanation of symbols

0400 Video processing device 0401 3DNR unit 0402 Judgment unit 0403 Noise reduction amount determination unit

Claims (3)

A 3DNR unit that performs processing to reduce noise superimposed on the video signal;
A determination unit that determines whether the acquired video source is a pull-down source;
When the determination result is a determination result indicating that the source is a pull-down source, a noise reduction amount determination unit that determines a noise reduction amount in the 3DNR unit according to the repetition rule of the same video frame in the video source;
A video processing apparatus comprising:
The noise reduction amount determination unit
A determination means for determining a pull-down source type;
When the determination result is 2: 3 pull-down, the noise reduction amount is set to the maximum threshold between the same video of the two consecutive identical videos or the three consecutive identical videos that appear, and the noise is reduced between the change videos across the same video group. First dynamic control means for determining the amount to a value according to the amount of image change;
A video processing apparatus. The noise reduction amount determination unit
A determination means for determining a pull-down source type;
When the determination result is 2: 2 pull-down, the noise reduction amount is set to the maximum threshold between the same images of the two consecutive identical videos that appear, and the noise reduction amount is changed between the changed videos across the same video group. Second dynamic control means for determining a value according to the amount;
The video processing apparatus according to claim 1, comprising: A determination step of determining whether the acquired video source is a pull-down source;
If the determination result is a determination result that is a pull-down source, a noise reduction amount determination step that determines a noise reduction amount according to the repetition rule of the same video frame in the video source;
A 3DNR step for performing a process of reducing noise superimposed on the video signal based on the noise reduction amount determined in the noise reduction amount determination step;
A noise reduction method by a video processing device having
The noise reduction amount determination step
A determination sub-step for determining a pull-down source type;
When the determination result is 2: 3 pull-down, the noise reduction amount is set to the maximum threshold between the same video of the two consecutive identical videos or the three consecutive identical videos that appear, and the noise is reduced between the change videos across the same video group. A first dynamic control substep for determining the amount to a value according to the amount of image change;
A noise reduction method.