JP5941660B2 - Noise detection device, playback device, and noise detection program - Google Patents

Description

  The present invention relates to a noise detection device, a playback device, and a noise detection program.

  In a moving image, it is known that inherent noise due to a specific factor occurs. For example, there is flashing noise in which luminance and color change greatly within a short time in a moving image. Therefore, it is considered to automate the detection of such blinking noise. For example, the invention of Patent Document 1 discloses a technique for automatically detecting blinking noise by setting a plurality of sample points representing sampling positions in a frame image and sequentially extracting and determining color information of the sample points. Has been.

Japanese Patent No. 4476125

  However, in the invention of Patent Document 1, since the color information of each sample point is averaged over the entire frame, correct detection may not be performed. For example, the average value obtained by the above-described averaging may be the same for “flashing noise with a part of the frame being very bright” and “flashing noise with a moderately bright part of the entire frame”.

  An object of the present invention is to provide means for automatically and accurately detecting blinking noise included in a moving image.

The noise detection apparatus according to one aspect includes an acquisition unit that acquires information of a moving image to be processed, a dividing unit that divides each of a plurality of images constituting the moving image into a plurality of regions, Based on the difference between the feature amount in the current image and the feature amount in the previous image between the plurality of images, a calculation unit that calculates a feature amount related to the blinking noise to be detected for each region A state determination unit that determines whether the change in the feature amount of the region is an increase or decrease, and a difference between the feature amounts in an increase period or a decrease period of the change in the feature amount of the region determined by the state determination unit of based on the cumulative value, and whether the determined area noise judgment unit or a region noise included in the area of the arbitrary of the moving image, based on the region the noise determination unit of the judgment result, the on the moving image Flashing noise A noise determination unit that determines whether or not it is rare, and the increase period is a period that is continuously determined to increase until it is determined to be decreased by the state determination unit, and the decrease period is an increase This is a period during which it is continuously determined that the value decreases until it is determined.

  The state determination unit may maintain the previous determination when the difference between the feature amount in the current image and the feature amount in the previous image does not exceed a predetermined threshold. .

The area noise determination unit may include the area noise in the area when a cumulative value of the difference between the feature quantities in an increase period or a decrease period of the change in the feature quantity of the area is larger than a predetermined threshold. You may judge.
In addition, a feature amount change amount calculation unit may be included, and the feature amount change amount calculation unit may accumulate the difference value when the determination by the state determination unit is the same as the determination in the previous image. .

  The state determination unit may determine that the state is in a steady state when either the increase or the decrease is maintained for a predetermined time.

Further, the noise determination unit determines whether or not the blinking noise is included in each image unit constituting the moving image based on the determination result by the region noise determination unit, and based on the determination result, It may be determined whether or not the blinking noise is included in the moving image .

In addition, the noise determination unit includes the blinking noise more than a predetermined number of times within a predetermined time in the moving image based on a result of determining whether the blinking noise is included in the image unit. Whether or not it may be determined.

  Further, the calculation unit may calculate at least one of a feature value related to luminance of the image and a feature value related to color information of the image as the feature value.

The noise detection device according to one aspect includes an acquisition unit that acquires information of a moving image including a first image and a second image, a plurality of regions of the first image, and a plurality of regions of the second image. A calculation unit that calculates information related to blinking noise in a region, and determines whether the difference between the information in the region of the first image and the information in the region of the second image is an increase or decrease. The information of the region of the first image and the second image in the first period, which is a period in which the first determination unit is continuously determined to increase until it is determined to be decreased by the first determination unit The area of the first image in the second period, which is a value obtained by accumulating the difference between the area and the information, and a period in which the first determination unit continuously determines a decrease until the increase is determined. And the information of the region of the second image Differences based on at least one of a value obtained by accumulating, and a second determination unit configured to determine whether include the flashing noise on the moving image.
Further, the image processing apparatus includes a dividing unit that divides the first image into a plurality of the regions and divides the second image into the plurality of regions.
The playback device according to one aspect includes a recording unit that records a moving image, a playback unit that plays back the moving image, and any one of the noise detection devices described above, and the acquisition unit receives the moving image from the recording unit. Get image information.

The noise detection program of one aspect includes an acquisition process for acquiring information on a moving image to be processed, a dividing process for dividing each of a plurality of images constituting the moving image into a plurality of regions, For each region, based on a calculation process for calculating a feature amount related to blinking noise to be detected, and a difference between the feature amount in the current image and the feature amount in the previous image between the plurality of images. A state determination process for determining whether the change in the feature amount of the region is an increase or a decrease, and a difference between the feature amounts in an increase period or a decrease period of a change in the feature amount of the region determined by the state determination process of based on the cumulative value, and determines the area noise determination process whether include regions noise in any area of the moving image, based on a determination result of the region noise judgment process, the A noise determination process it is determined whether or not the included flashing noise in the image, which causes a computer to execute the said increase periods, and continuously increased until it is determined that the reduction by the state determination process It is a determined period, and the decrease period is a period determined to decrease continuously until it is determined to increase by the state determination process.
The noise detection program according to one aspect includes an acquisition process for acquiring information on a moving image having a first image and a second image, a plurality of regions of the first image, and a plurality of regions of the second image. A calculation process for calculating information related to blinking noise in a region, and determining whether the difference between the information in the region of the first image and the information in the region of the second image is an increase or decrease. The first determination process and the information of the region of the first image and the second image in the first period, which is a period determined to be increased continuously until it is determined to be decreased by the first determination process The area of the first image in the second period, which is a value obtained by accumulating the difference between the information of the area and the decrease continuously until the increase is determined by the first determination process The information and the second image Based the difference between the information of the frequency band on at least one value obtained by accumulating, a second determination process of determining whether or not include the flash noise on the moving image, causes the computer to execute.
In this case, the image processing apparatus includes a dividing process of dividing the first image into the plurality of areas and dividing the second image into the plurality of areas.

  According to the present invention, it is possible to provide means for automatically and accurately detecting blinking noise included in a moving image.

Block diagram showing a configuration example of a noise detection device Flow chart showing an example of operation of the noise detection device during noise detection Flow chart showing an example of the operation of the noise detection device during noise detection (continued) Flow chart showing an example of the operation of the noise detection device during noise detection (continued) Diagram explaining split processing Example of weighting table for hue Example of weighting table for saturation Example of a weighting table for brightness The figure explaining the change state of a difference value Another diagram explaining the change state of the difference value The figure explaining the modification of a brightness | luminance blink detection

  FIG. 1 is a block diagram illustrating a configuration example of a noise detection device according to an embodiment. In the noise detection apparatus, a noise detection program for detecting noise is installed in advance for a moving image (target moving image) to be processed.

  Note that any target moving image may be used. For example, it may be generated by an imaging device capable of generating a digital image, may be converted from analog moving image image data to digital data, or recorded on a video tape or the like. A moving image may be converted into digital data. Further, it may be a moving image created by a computer or the like.

  In the following, an example in which all the frames of the target moving image are processed as noise detection targets will be described. However, only a part of the frames of the target moving image may be processed as noise detection targets. In this case, the noise detection target frame may be specified based on a user operation, or may be automatically specified based on information on the target moving image.

  A noise detection device 11 shown in FIG. 1 includes a noise detection unit 12, a CPU 13, an alarm output unit 14, an input / output I / F 15, and a bus 16. The CPU 13 is connected to the input / output I / F 14. The output of the noise detection unit 12 is connected to the CPU 13 via the bus 16. Further, the alarm output unit 14 is controlled by the CPU 13.

  The image data of the target moving image input to the noise detection device 11 is input to the noise detection unit 12.

  Furthermore, an input device (such as a keyboard and pointing device) (not shown), a monitor, a network cable for LAN connection, and the like are connected to the noise detection device 11 via an input / output I / F 14.

  The noise detection device 11 includes a storage device (not shown) (for example, a storage medium such as a hard disk or a nonvolatile semiconductor memory). This storage device stores an image processing program and various data necessary for executing the program.

  The CPU 13 is a processor that comprehensively controls each unit of the noise detection device 11. The alarm output unit 14 includes a notification member such as a speaker, a monitor, and a lamp, and notifies the user of the detection result by the noise detection unit 12.

  As shown in FIG. 1, the noise detection unit 12 includes a YHSV conversion unit 31 to which a target moving image is input, a luminance blinking noise detection unit 22 that detects luminance blinking noise, and a color blinking noise detection unit 23 that detects color blinking noise. Have The luminance flashing noise detection unit 22 includes a Y evaluation value calculation unit 31, a block luminance flashing detection unit 32, a frame luminance flashing determination unit 33, and a luminance flashing determination unit 34. Each unit includes an evaluation value calculation unit 35, an S evaluation value calculation unit 36, a V evaluation value calculation unit 37, a block color information calculation unit 38, a block color blink detection unit 39, a frame color blink determination unit 40, and a color blink determination unit 41. .

  The noise detection device 11 having the above-described configuration detects predetermined inherent blinking noise. The flashing noise to be detected may be anything, but in the following, “bright red flashing” is “frequency more than a predetermined area of one frame” and “three times per second” The above-described flashing noise will be described as an example of a detection target. In the following, an example in which two elements of “brightness blinking” and “color blinking” are detected when the above-described blinking noise is detected will be described. It should be noted that “inherent blinking noise” to be detected includes blinking that exists in the target moving image based on the need for expression. For example, blinking for expressing an explosion or the like in an animation can be a detection target of the noise detection device 11 as “inherent blinking noise”.

  First, an operation example of blinking noise detection in the noise detection unit 12 will be described with reference to the flowcharts of FIGS. The processing in the flowcharts of FIGS. 2 to 4 is started when each unit executes a noise detection program in accordance with a program execution instruction from the user.

(Step S101)
The noise detection unit 12 acquires one frame of image data of the target moving image specified by the user. The acquired image data of one frame is input to the YHSV conversion unit 21 as shown in FIG.

(Step S102)
The noise detection unit 12 performs YHSV conversion processing by the YHSV conversion unit 21.

  The YHSV conversion unit 21 performs a known YHSV conversion process on the input signal to obtain a luminance signal (hereinafter referred to as a Y signal), a hue signal (hereinafter referred to as an H signal), and a saturation signal (hereinafter referred to as an S signal). And lightness signals (hereinafter referred to as V signals). The contents of the YHSV conversion process vary depending on the format of the input signal. For example, when the input signal is a YCbCr signal, the Y component is output as it is as a Y signal, and an H signal, an S signal, and a V signal are generated. For example, when the input signal is an RGB signal, a conversion process from a known RGB color space to a YHSV color space is performed to generate a Y signal, an H signal, an S signal, and a V signal.

  The generated Y signal is input to the Y evaluation value calculation unit 31. Further, the generated H signal is input to the H evaluation value calculation unit 35. Further, the generated S signal is input to the S evaluation value calculation unit 36. Further, the generated V signal is input to the V evaluation value calculation unit 37.

(Step S103)
The noise detection unit 12 performs a division process using the Y evaluation value calculation unit 31. The Y evaluation value calculation unit 31 divides the input Y signal into a plurality of regions for each frame. For example, as shown in FIG. 5, the Y evaluation value calculation unit 31 divides the Y signal of each frame into 25 regions of 5 in each of the vertical and horizontal directions. The 25 areas resulting from this division are referred to as B (1) to B (25), respectively.

  In addition, the noise detection unit 12 performs the same division processing on the H signal, the S signal, and the V signal by the H evaluation value calculation unit 35, the S evaluation value calculation unit 36, and the V evaluation value calculation unit 37, respectively. The area is divided into 25 areas B (1) to B (25).

  In the example of FIG. 5, an example of equal division into 5 × 5 is shown, but other division methods may be used. For example, it may be divided into regions of different sizes depending on the location in the frame.

  Further, the division method may be appropriately changed according to the content of the target moving image, the shooting mode when shooting the target moving image, and the like.

(Step S104)
The noise detection unit 12 performs an evaluation value calculation process using the Y evaluation value calculation unit 31. The Y evaluation value calculation unit 31 calculates an evaluation value for each of 25 areas B (1) to B (25) divided in step S103. For example, the Y evaluation value calculation unit 31 calculates an average value of Y signals included in each region as an evaluation value. As a result, 25 evaluation values (hereinafter referred to as Ey (1) to Ey (25)) are calculated corresponding to 25 regions of B (1) to B (25). This evaluation value is input to the block luminance blink detection unit 32 and is used to detect luminance blink noise (details will be described later).

  Further, the noise detection unit 12 performs the same processing on the H signal by the H evaluation value calculation unit 35 to calculate 25 evaluation values (hereinafter referred to as Eh (1) to Eh (25)). The S evaluation value calculation unit 36 performs the same processing on the S signal to calculate 25 evaluation values (hereinafter referred to as Es (1) to Es (25)), and a V evaluation value calculation unit 37. Thus, the same processing is performed on the V signal, and 25 evaluation values (hereinafter referred to as Ev (1) to Ev (25)) are calculated.

  Further, the noise detection unit 12 performs an evaluation value calculation process on the color information by the block color information calculation unit 38. The block color information calculation unit 38 includes the evaluation values based on the H signal (Eh (1) to Eh (25)), the evaluation values based on the S signal (Es (1) to Es (25)), and the V signal. The evaluation values related to color information (Ev (1) to Ev (25)) are weighted and integrated to calculate evaluation values relating to color information (hereinafter referred to as Ec (1) to Ec (25)).

  The weight is determined according to the nature of the flashing noise to be detected. In this embodiment, since the flashing noise of “bright red flashing” is a detection target, weighting is performed using a table in which the hue with the highest red purity reaches its peak.

  6 to 8 are examples of weighting tables (referred to as Th, Ts, and Tv, respectively) relating to hue, saturation, and brightness. As shown in FIG. 6, with respect to the saturation, a table is used such that the saturation reaches 0 ° in accordance with the blinking noise of “bright red blinking”.

  The block color information calculation unit 38 calculates an evaluation value related to color information (hereinafter referred to as Ec (1) to Ec (25)) using these tables for each region.

  The evaluation value Ec (X) in an arbitrary region B (X) is calculated by the following equation.

Ec (X) = {Eh (X)} weighted by Th} * {Es (X)} weighted by Ts} * {Ev (X)} weighted by Tv}
The block color information calculation unit 38 performs the same processing for each of B (1) to B (25), and calculates evaluation values (Ec (1) to Ec (25)) related to the color information. This evaluation value is input to the block color blink detection unit 39 and used for detection of color blink noise (details will be described later).

  In addition, when performing weighting, you may use methods other than a table. In addition, the saturation-related table Ts and the brightness-related table Tv may be prepared, or the same table may be used.

  Further, when calculating the evaluation value related to the luminance information and the evaluation value related to the color information, the evaluation value may be calculated after appropriately weighting each area. For example, the evaluation value may be calculated after the weight is increased for the central region (for example, the region B (13) in FIG. 5) and weighted for each region so that the weight decreases toward the outside. .

(Step S105)
The noise detection unit 12 performs luminance blink detection by each of the block luminance blink detection unit 32, the frame luminance blink determination unit 33, and the luminance blink determination unit 34. Details of the detection will be described later.

(Step S106)
The noise detection unit 12 performs color blink detection by the block color blink detection unit 39, the frame color blink determination unit 40, and the color blink determination unit 41. Details of the detection will be described later.

(Step S107)
The CPU 13 determines whether or not the processing from step S102 to step S106 has been performed for all frames. If the CPU 13 determines that the process has been completed for all frames, the CPU 13 ends the series of processes. If the CPU 13 determines that the process has not been completed for all frames, the process proceeds to step S108.

(Step S108)
The noise detection unit 12 acquires the image data of the next frame and returns to step S102.

  Next, an example of the operation of detecting the luminance blinking noise in the luminance blinking noise detection unit 22 during the luminance blinking detection process in step S105 will be described with reference to the flowchart of FIG.

(Step S121)
The noise detection unit 12 uses the block luminance blink detection unit 32 to detect luminance blink in units of blocks. The block luminance blinking detection unit 32 detects luminance blinking in units of blocks by performing the following four processes for each region. By this detection, it is possible to detect an area (block) including luminance blinking among the areas B (1) to B (25). In the following, an explanation will be given by taking an arbitrary region B (X) as an example.
(A) Calculation of Difference Value Each time a frame image is acquired, the block luminance blink detection unit 32 evaluates the evaluation value Ey (X regarding luminance information calculated in step S104 for the acquired frame and the previous frame. ) Is calculated.
(B) State Monitoring The block luminance blink detection unit 32 is in any state of “increase”, “decreasing”, and “steady” as described above, according to the condition shown in FIG. It is monitored whether it is. In FIG. 9, “with effective increase” means that the above-described difference value is a positive value (evaluated value of the acquired frame> evaluated value of the previous frame), and the absolute value of the difference value is a predetermined value. This is the case when it is larger than the threshold value. In such a case, the state becomes “increase”. In FIG. 9, “with effective reduction” means that the above-described difference value is a negative value (evaluation value of the acquired frame <evaluation value of the previous frame), and the absolute value of the difference value is This is the case when it is larger than the predetermined threshold. In such a case, the state becomes “decreasing”. On the other hand, if the state is “no effective increase” and “no effective decrease” between predetermined frames, the state is “steady”. Note that the threshold for determining effective increase and decrease and the number of the predetermined frames described above are determined in advance according to the nature of the blinking noise to be detected.

  The changes in the three states described above will be exemplified with reference to FIG. FIG. 10 is an example showing a change in the evaluation value Ey (X) in an arbitrary region B (X). In FIG. 10, the dotted line graph shows the change in the analog signal of the evaluation value Ey (X), and the solid line graph shows the digital change in the evaluation value Ey (X) at each processing. Further, the horizontal axis of FIG. 10 indicates the frame number (that is, time), and the vertical axis indicates the evaluation value Ey (X). In FIG. 10, F (X) indicates that the frame number is X. In the example of FIG. 10, for the purpose of explanation, the frame numbers are assigned in order from “1”, but the same applies to any part of the target moving image. In the example of FIG. 10, only “in increasing” and “decreasing” among the above-described three states are illustrated, and “steady state” is not particularly illustrated.

  In FIG. 10, the change from F (2) to F (3) does not correspond to the above “effective increase”. This is because the evaluation value of F (3)> the evaluation value of F (2), but the absolute value of the difference value between the evaluation values of F (2) and F (3) is smaller than a predetermined threshold value. The same applies to the change from F (3) to F (4). On the other hand, the change from F (4) to F (5) corresponds to the above “effective increase”. This is because the evaluation value of F (5)> the evaluation value of F (4), and the absolute value of the difference value between the evaluation values of F (4) and F (5) is larger than a predetermined threshold value. . At this timing, the block luminance blink detection unit 32 determines that the state has changed to “increasing”, and executes “processing when the state is increasing” described later.

  Subsequently, since the change from F (5) to F (6) and the change from F (6) to F (7) do not satisfy any of the conditions described in FIG. 9, the state does not change. , Remains “increase”. The change from F (7) to F (8) corresponds to the “effective reduction” described above. This is because the evaluation value of F (8) <the evaluation value of F (7), and the absolute value of the difference value between the evaluation values of F (7) and F (8) is greater than a predetermined threshold value. . At this timing, the block luminance blink detection unit 32 determines that the state has changed to “decreasing”, and executes “processing when the state is decreasing” described later.

  Subsequently, since the change from F (8) to F (9) and the change from F (9) to F (10) do not satisfy any of the conditions described in FIG. 9, the state does not change. , Remains “decreasing”. Since the change from F (10) to F (11) corresponds to the “effective increase” described above, the block luminance blink detection unit 32 changes the state to “increase” at this timing. And “process when the state is increasing”, which will be described later, is executed.

The same applies to the following, but the change from F (15) to F (16) whose state is “increase” is F (16) evaluation value <F (15) evaluation value, but F (15 ) And F (16), the absolute value of the difference value between the evaluation values is smaller than the predetermined threshold value, and therefore does not correspond to the above “effective reduction”, but becomes “invalid reduction”. Therefore, the state does not change and remains “increase”.
(C) Processing by state <Increase>
When the state is “increasing”, the block luminance blinking detection unit 32 starts with “effective increase” when the state changes while increasing, in addition to the above-described calculation of the difference value and monitoring of the state. The accumulated difference values are accumulated and updated as accumulated values. For example, in the example of FIG. 10, since the state changes to “increase” at the timing of change from F (4) to F (5), the difference value is accumulated from F (4) as shown in FIG. Will be. The cumulative value is updated until the state changes from “increase” to another state. The accumulated value obtained in this way is used when detecting a blinking candidate described later.
<Decrease>
When the state is “decreasing”, the block luminance blink detection unit 32 performs only the above-described difference value calculation and state monitoring. Note that the block luminance blinking detection unit 32 may perform the above-described calculation of the cumulative value even during a decrease, and may take this into account when a blinking candidate described later is detected.
<Stationary>
When the state is “steady”, the block luminance blink detection unit 32 performs only the difference value calculation and the state monitoring described above.
(D) Detection process when the state changes from “Increased” to “Decreased” to any other state The state changes from “Increased” to “Decreased” to any other state Then, the block luminance blink detection unit 32 first compares the accumulated value held in the above “increase” with a predetermined threshold. The other arbitrary states described above include a case where the difference value becomes 0 or more in addition to the state of “increase” or “steady state”. If cumulative value> predetermined threshold value, it is estimated that there is a blinking candidate in the vicinity of the frame corresponding to the starting point of “effective decrease” when the state changes to “decreasing”, and the brightness at this time It is determined that blinking noise is included. For example, in the example of FIG. 10, the state changes to “increase” according to the change from F (4) to F (5), and further according to the change from F (7) to F (8). Since the state changes to “decreasing”, and the state changes to “increasing” in accordance with the change from F (10) to F (11), the state changes to “decreasing”. The starting point of “effective reduction” is F (7), and it is estimated that there is a blinking candidate in the vicinity of F (7), and it is determined that luminance blinking noise is included at this time.

  Note that the case where the accumulated value <the predetermined threshold value is a case where the touch width of the evaluation value related to luminance is small and blinking insufficient to determine luminance blinking noise is included. The threshold value to be compared with the accumulated value is determined in advance according to the nature of the flashing noise to be detected.

  In addition, “steady” is included in the state change when the touch range of the evaluation value related to the brightness is sufficient, while the change time is long and the change rate of the evaluation value related to the brightness includes flashing slowly. is there. Such blinking is not detected as blinking noise although there is an extreme value of the evaluation value.

  The block luminance blinking detection unit 32 performs the above-described four processes for each of the regions B (1) to B (25), and proceeds to step S122.

  It should be noted that the method for detecting luminance blinking described in step S121 is an example. For example, luminance blinking may be detected by simply obtaining a maximum value or a minimum value. Further, for example, the luminance blinking may be detected by detecting only the rising portion of the evaluation value change.

  In addition, in the detection of the luminance blink described in step S121, the luminance blink may be detected in consideration of the brightness of the base in the change in the evaluation value. For example, the darker the brightness of the base, the greater the influence of the blinking luminance, and the brighter the base, the less noticeable the influence of the blinking luminance. Therefore, the brightness blinking may be detected in consideration of the brightness of the base in the change in the evaluation value, for example, the threshold value to be compared with the cumulative value is reduced as the brightness of the base is darker.

(Step S122)
The noise detection unit 12 determines whether or not there is luminance blinking in units of blocks based on the result of detection of luminance blinking in units of blocks performed in step S121. If the noise detection unit 12 determines that there is a luminance blink in units of blocks, the process proceeds to step S123. If the noise detection unit 12 determines that there is no luminance blink in units of blocks, the series of processing ends.

(Step S123)
In the noise detection unit 12, the frame luminance blink determination unit 33 detects luminance blink in units of frames. The frame luminance blink determination unit 33 counts the number of areas in which the luminance blink is detected in the frame, and when the number of areas is larger than a predetermined threshold, the frame has a luminance blink noise. It is determined that it is included. The threshold value is determined according to the nature of the blinking noise to be detected. In the present embodiment, since the flashing noise of “frequency more than a predetermined area of one frame” is a detection target, for example, when the threshold is 5, the number of areas in which luminance flashing is detected is 5 or more, That is, when luminance blinking occurs in an area of 20% or more of a frame, it is determined that luminance blinking noise is included in the frame.

  As described above, by determining whether or not blinking noise is included by setting a threshold value in the area, it is possible to detect both blinking noise that occurs discretely and blinking noise that occurs collectively.

(Step S124)
The noise detection unit 12 determines whether or not there is a luminance blink in the frame unit based on the result of the luminance blink detection in the frame unit performed in step S123. If the noise detection unit 12 determines that there is a luminance blink in units of frames, the process proceeds to step S125. If the noise detection unit 12 determines that there is no luminance blink in units of frames, the series of processing ends.

(Step S125)
The noise detection unit 12 determines whether or not the number of blinks within a predetermined time is greater than or equal to the threshold value Tt for the luminance blink in units of frames by the luminance blink determination unit 34. If the noise detection unit 12 determines that the number of blinks within the predetermined time is equal to or greater than the threshold value Tt, the process proceeds to step S126. If the noise detection unit 12 determines that the number of blinks within the predetermined time is less than the threshold value Tt, the series of processing ends. .

  The threshold value Th is determined according to the nature of the blinking noise to be detected. In the present embodiment, since the flashing noise of “three times or more per second” is a detection target, the threshold Th of the number of flashes per second becomes Th = 3.

(Step S126)
The noise detection unit 12 outputs information (occurrence frame No., generation time, etc.) indicating that the luminance blinking noise exists as determination contents to the CPU 13 via the bus 16 and ends the series of processes.

  By performing such processing, for example, when the blinking of the luminance in units of frames occurs once every 3 seconds, it is determined that there is no luminance blinking noise.

  Next, an operation example of color blinking noise detection in the color blinking noise detection unit 23 during the color blinking detection process in step S106 will be described with reference to the flowchart of FIG. Note that the detection of the color blinking noise includes a process common to the detection of the luminance blinking noise described with reference to FIG. 3, and therefore only the parts different from the detection of the luminance blinking noise will be described below.

(Step S131)
The noise detection unit 12 uses the evaluation values (Ec (1) to Ec (25)) relating to the color information calculated in step S104 by the block color blinking detection unit 39, and performs the same processing as in step S121 in FIG. To detect color blinking in block units.

(Step S132)
The noise detection unit 12 determines whether or not there is color blinking in units of blocks based on the result of detection of color blinking in units of blocks performed in step S131. If the noise detection unit 12 determines that there is color blinking in units of blocks, the process proceeds to step S133. If the noise detection unit 12 determines that there is no color blinking in units of blocks, the series of processing ends.

(Step S133)
The noise detection unit 12 uses the frame color blink determination unit 40 to perform the same process as in step S123 of FIG. 3 described above, and detects the color blink in units of frames.

(Step S134)
The noise detection unit 12 determines whether or not there is color blinking in units of frames based on the result of detection of color blinking in units of frames performed in step S133. If the noise detection unit 12 determines that there is color blinking in units of frames, the process proceeds to step S135, and if it is determined that there is no color blinking in units of frames, the series of processing ends.

(Step S135)
The noise detection unit 12 uses the color blink determination unit 41 to determine whether or not the number of blinks within a predetermined time is greater than or equal to the threshold value Tt for the color blink in units of frames. If the noise detection unit 12 determines that the number of blinks within the predetermined time is greater than or equal to the threshold value Tt, the process proceeds to step S136. If the noise detection unit 12 determines that the number of blinks within the predetermined time is less than the threshold value Tt, the series of processing ends. .

(Step S136)
The noise detection unit 12 outputs information (occurrence frame No., generation time, etc.) indicating that the color blinking noise exists as the determination content to the CPU 13 via the bus 16 and ends the series of processes.

  In addition, each threshold value used for the detection of the color blinking noise may be the same as or different from each threshold value used for the detection of the luminance blinking noise described with reference to FIG.

  Further, in the above-described example, an example in which the luminance blinking noise detection and the color blinking noise detection are performed independently is shown. However, even if noise detection is performed by judging both detection results in combination. good. One detection result may be reflected in the other detection.

  In the above example, a series of detection processes are performed every time one frame of image data is acquired. However, a series of detection processes may be performed every time several frames of image data are acquired. A series of detection processes may be performed after all frames have been acquired. In this case, for example, more detailed flashing noise can be detected by creating the number of blocks of the difference value change state shown in FIG. 10 and analyzing these graphs comprehensively. .

Further, for example, even when the shape of the blinking noise slightly changes between successive frames, the process shown in FIG. 11 may be further performed in order to reliably detect the blinking noise. The step numbers in FIG. 11 correspond to the step numbers in FIG. As shown in FIG. 11, when the blinking noise is slightly moved between F (n−1) and F (n) as a result of the detection of the luminance blinking in units of blocks in step S121, the step is performed. When the luminance blinking is detected in units of frames in S123, the number of areas in which the luminance blinking is detected is 5 or less, and therefore the luminance blinking is not detected in both F (n−1) and F (n). . Therefore, as shown in FIG. 11, as a step S121 ^* , a logical sum (OR synthesis) with luminance blinking in units of blocks in the previous frame is performed. By performing such processing, for example, the number of areas in which luminance blinking is detected in F (n) becomes five or more, and therefore, luminance blinking in units of frames can be detected in F (n). Note that the blinking noise that is the detection target in this embodiment is not included in two consecutive frames. Therefore, when the above-described processing is performed, if the blinking of brightness in units of frames is detected for two consecutive frames, the blinking of the brightness of one of the frames is ignored, and only the blinking of the brightness of one frame is valid. Good.

  The noise detection result may be used for notification to the user via the alarm output unit 14, or may be output to the outside of the noise detection device 11 via the input / output I / F1. Further, information indicating the type of detected noise may be recorded in the tag information of the target moving image.

  Moreover, each process demonstrated in the above-mentioned example may be performed by hardware, and may be performed by the software by CPU13.

  As described above, the noise detection apparatus according to the present embodiment acquires information on a target moving image to be processed, and divides each of a plurality of images constituting the target moving image into a plurality of regions. And the feature-value regarding the blink noise of detection object is calculated for every area | region. Furthermore, it is determined whether or not the change in the feature amount of the region is an increase or decrease based on the difference between the feature amount in the current image and the feature amount in the previous image between a plurality of images. Based on the state, the change amount of the feature amount is calculated. Then, when the determined state changes from increase to decrease and changes to any other state, it is determined whether or not the region noise is included in any region of the moving image based on the change amount of the feature amount. Based on the determination result, it is determined whether or not the moving image includes blinking noise.

  Therefore, according to the configuration of the present embodiment, even when the blinking noise included in the moving image is locally included in the image, the blinking noise can be detected. Moreover, it is possible to prevent erroneous detection and omission of detection by determining the state based on the change over time and detecting the flashing noise according to the state. Therefore, blinking noise included in the moving image can be automatically detected accurately and accurately.

  Further, according to the present embodiment, it is determined whether or not blinking noise is included in each image constituting the moving image, and whether or not the target moving image includes blinking noise is determined based on the determination result. Make a decision. Therefore, it is possible to distinguish and detect both the flashing noise generated locally and the flashing noise generated overall.

  In the above-described example, an example in which a series of processing is executed in accordance with a program execution instruction by the user has been shown, but the present invention is not limited to this example. For example, a series of processing may be automatically executed every time the noise detection device 11 reads image data of some moving image.

<Supplementary items of the embodiment>
(1) Each process, each threshold value, etc. which were demonstrated by the said embodiment are examples, and this invention is not limited to this example.

  For example, in the above embodiment, “brilliant red blinking” is “frequently more than a predetermined area of one frame” and “flashing noise more than three times per second” is cited as a detection target. The flashing noise having the above characteristics may be detected. For example, “cyan blinking” may be detected. When flashing noise other than red is to be detected, various tables and threshold values may be determined according to the nature of the noise. Moreover, it is good also considering the color blink noise of multiple colors as a detection target. In such a case, the color blinking noise of each color may be detected independently.

  Moreover, although the example which performs a series of processes after converting color space into HSV color space by YHSV conversion part 21 was shown in the said embodiment, it is good also as a structure converted into color spaces other than HSV color space.

  In the above-described embodiment, an example in which blinking noise is detected in order of a block unit, a frame unit, and a moving image unit is shown. However, when a moving image to be detected includes a plurality of fields, a block unit and a field unit are used. The flashing noise may be detected in the order of moving image units. Alternatively, the detection of blinking noise may be omitted in units of frames, and it may be determined whether or not blinking noise is included in the entire moving image based on the result of detecting blinking noise in units of blocks.

  (2) A noise detection program for executing processing by the noise detection apparatus described in the above embodiment is also effective as a specific aspect of the present invention. This noise detection program may be stored in a storage medium such as a magnetic disk, or may be downloaded via the Internet or the like. Further, an image processing apparatus and a reproduction apparatus (for example, a photo viewer, a digital photo frame, various printing apparatuses, etc.) provided with the noise detection apparatus described in the above embodiment are also effective as specific embodiments of the present invention. The playback apparatus described above may be configured to execute a series of processes when playing back an image.

  (3) Each image processing and each determination method of the above embodiment is an example, and the present invention is not limited to this example.

  From the above detailed description, features and advantages of the embodiments will become apparent. It is intended that the scope of the claims extend to the features and advantages of the embodiments as described above without departing from the spirit and scope of the right. Further, any person having ordinary knowledge in the technical field should be able to easily come up with any improvements and modifications, and there is no intention to limit the scope of the embodiments having the invention to those described above. It is also possible to use appropriate improvements and equivalents within the scope disclosed in.

DESCRIPTION OF SYMBOLS 11 ... Noise detection apparatus, 12 ... Noise detection part, 13 ... CPU

Claims (14)

An acquisition unit for acquiring information of a moving image to be processed;
A dividing unit that divides each of a plurality of images constituting the moving image into a plurality of regions;
For each region, a calculation unit that calculates a feature amount related to the blinking noise to be detected;
It is determined whether the change in the feature amount of the region is an increase or decrease based on a difference between the feature amount in the current image and the feature amount in the previous image between the plurality of images. A state determination unit;
Based on the cumulative value of the difference of the feature quantity at an increase period or a decrease period of a change in the characteristic quantity of the region determined by the state determination unit, whether or not included regions noise region of arbitrary of the moving image An area noise determination unit for determining
A noise determination unit that determines whether or not the blinking noise is included in the moving image based on a determination result of the region noise determination unit ;
The increase period is a period continuously determined to increase until it is determined to be decreased by the state determination unit,
The noise detecting device is a period in which the decrease period is continuously determined to decrease until the state determination unit determines to increase . The noise detection device according to claim 1,
The state determination unit, when the difference between the feature amount in the characteristic amount and the previous image in the current image does not exceed a predetermined threshold value, noise detection apparatus that maintain the previous determination . In the noise detection device according to claim 1 or 2,
The region noise determining unit determines that the region includes the region noise when a cumulative value of the difference between the feature amounts during an increase period or a decrease period of the feature amount change of the region is larger than a predetermined threshold. that noise detection device. In the noise detection device according to any one of claims 1 to 3,
It has a feature amount change amount calculation unit,
The feature quantity variation calculation unit may, when the determination by the state determining unit of the same determination in the previous image noise detecting device it accumulate values of the difference. In the noise detection device according to any one of claims 1 to 4,
Noise detection device the condition judging unit you determined steady state when maintained either state of the increase or the decrease predetermined time. In the noise detection apparatus according to any one of claims 1 to 5,
Wherein the noise determination unit, based on a determination result of the region the noise determination unit determines whether or include the flashing noise in each image units constituting the moving image, based on the determination result, the videos It is determined whether or not the included flashing noise in an image line Uno size detecting device. In the noise detection device according to any one of claims 1 to 6,
Whether the noise determination unit includes the blinking noise more than a predetermined number of times within a predetermined time in the moving image based on a determination result of determining whether the blinking noise is included in the image unit. whether the determined row Uno size detecting device. In the noise detection device according to any one of claims 1 to 7,
The calculating unit, as the feature quantity, the feature quantity relating to the luminance of the image, noise detection apparatus you calculate at least one of the feature amount relating to the color information of the image.   An acquisition unit for acquiring information of a moving image having a first image and a second image;
  A calculation unit that calculates information regarding blinking noise of the plurality of regions of the first image and the plurality of regions of the second image;
  A first determination unit that determines whether the difference between the information of the region of the first image and the information of the region of the second image is an increase or a decrease;
  The information of the area of the first image and the information of the area of the second image in a first period, which is a period continuously determined to increase until it is determined to be decreased by the first determination unit And the information of the region of the first image in the second period, which is a period in which the difference is accumulated until it is determined to be increased by the first determination unit, and the first A noise detection apparatus comprising: a second determination unit configured to determine whether or not the blinking noise is included in the moving image based on at least one of values obtained by accumulating a difference from the information of the region of two images. The noise detection device according to claim 9,
  A noise detection apparatus comprising: a dividing unit that divides the first image into a plurality of the regions and divides the second image into the plurality of regions.   A recording unit for recording moving images;
  A playback unit for playing back the moving image;
  A noise detection device according to any one of claims 1 to 10,
  The acquisition unit is a playback device that acquires information of the moving image from the recording unit.   An acquisition process for acquiring information of a moving image to be processed;
  A dividing process of dividing each of the plurality of images constituting the moving image into a plurality of regions;
  For each region, a calculation process for calculating a feature amount related to the blinking noise to be detected;
  It is determined whether the change in the feature amount of the region is an increase or decrease based on a difference between the feature amount in the current image and the feature amount in the previous image between the plurality of images. State determination processing;
  Whether or not a region noise is included in an arbitrary region of the moving image based on a cumulative value of the difference between the feature amounts in an increase period or a decrease period of the feature amount change of the region determined by the state determination process. An area noise determination process for determining;
  A noise determination process for determining whether or not the blinking noise is included in the moving image based on a determination result of the area noise determination process;
  To run on a computer,
  The increase period is a period continuously determined to increase until it is determined to be decreased by the state determination process,
  The noise detection program, wherein the decrease period is a period continuously determined to decrease until it is determined to be increased by the state determination process.   An acquisition process for acquiring information of a moving image having a first image and a second image;
  A calculation process for calculating information related to blinking noise of the plurality of regions of the first image and the plurality of regions of the second image;
  A first determination process for determining whether a difference between the information of the region of the first image and the information of the region of the second image is an increase or a decrease;
  The information of the area of the first image and the information of the area of the second image in a first period, which is a period determined to increase continuously until it is determined to decrease by the first determination process And the information of the region of the first image in the second period, which is a period in which the difference is accumulated until it is determined to be increased by the first determination process, and the first A second determination process for determining whether or not the moving image includes the blinking noise based on at least one of values obtained by accumulating a difference between the information of the two images and the area;
  A noise detection program that runs a computer. In the noise detection program according to claim 13,
  The noise detection program which has a division | segmentation process which divides | segments the said 1st image into the said some area | region, and divides | segments the said 2nd image into the said some area | region.

Images