JP6199749B2 - Noise judgment device and noise judgment method - Google Patents

Description

  The present invention relates to a noise determination technique, and more particularly, to a determination technique for noise included in an image signal.

  Conventionally, various techniques have been proposed for detecting and removing noise contained in an image signal captured using an image sensor such as a solid-state image sensor.

  For example, Patent Document 1 discloses a technique for detecting and removing noise included in a chroma component.

Japanese Patent No. 5105215

  An imaging apparatus using an image sensor such as a solid-state imaging device may be used as a surveillance camera. However, the surveillance camera is not limited to use in a bright place, and may be used in a place with low illuminance. .

  In such a case, the noise ratio becomes high, and conventional noise detection techniques including Patent Document 1 cannot perform effective noise determination, and there is a problem that sufficient noise removal cannot be performed.

  The present invention has been made to solve the above-described problems, and an object thereof is to provide a noise determination technique capable of performing noise determination on an image captured under any circumstances. And

  A first aspect of the noise determination apparatus according to the present invention is a noise determination apparatus that determines the presence or absence of noise in image data given in units of frames, and frequency-decomposes the image data, and regarding the obtained frequency components, A predetermined number of data is sampled for each of a low frequency component having a relatively low frequency and a high frequency component having a relatively high frequency, and the presence or absence of noise is determined based on the ratio of the high frequency data to the low frequency data.

  According to a second aspect of the noise determination device of the present invention, the noise determination device divides the image data of one frame into a matrix having a predetermined number of pixels and performs frequency decomposition as a data block; For the frequency components obtained by the frequency resolving unit, a predetermined number of frequency components are sampled on each of the low frequency component side and the high frequency component side of the data block to obtain the low frequency side data and the high frequency side data. A frequency analysis unit that analyzes the presence or absence of noise based on whether a ratio of the high-frequency data to the low-frequency data is greater than a predetermined first threshold value.

  A third aspect of the noise determination device according to the present invention is such that when the number of the data blocks analyzed by the frequency analysis unit as having noise is larger than a predetermined threshold, A single frame noise determination unit that determines that one frame has noise is provided.

  In a fourth aspect of the noise determination apparatus according to the present invention, when the noise determination apparatus satisfies a predetermined determination condition when a plurality of frames determined to have noise in the single frame noise determination unit, A multi-frame noise determination unit that determines that a moving image has noise is provided.

  According to a fifth aspect of the noise determination apparatus of the present invention, the noise determination apparatus divides the image data of one frame into a matrix of a predetermined number of pixels and performs frequency decomposition as a data block; For the frequency components obtained by the frequency resolving unit, a predetermined number of frequency components are sampled on each of the low frequency component side and the high frequency component side of the data block to obtain the low frequency side data and the high frequency side data. Creating a data aggregation matrix in which data is added for each corresponding frequency component for all the data blocks constituting the one frame, and the high frequency side data is compared with the low frequency side data in the data aggregation matrix. The presence / absence of noise is analyzed based on whether the ratio is greater than a predetermined first threshold. It is provided with a wave number analysis unit.

  According to a sixth aspect of the noise determination apparatus of the present invention, when the plurality of frames determined as having noise by the frequency analysis unit satisfy a predetermined determination condition, the noise determination apparatus is a moving image. Are provided with a multi-frame noise determination unit that determines that the signal has noise.

  In a seventh aspect of the noise determination device according to the present invention, the determination condition is defined by a ratio of frames having noise among a plurality of frames constituting the moving image.

  In an eighth aspect of the noise determination apparatus according to the present invention, the determination condition is defined by a continuous number of frames having noise among a plurality of frames constituting the moving image.

  In a ninth aspect of the noise determination apparatus according to the present invention, the frequency decomposition unit performs frequency decomposition by discrete cosine transform, and the frequency analysis unit obtains a transform coefficient obtained by discrete cosine transform in the frequency decomposition unit. The AC component is normalized by dividing the AC component by the DC component, and the AC component after normalization is used as the low-frequency data and the high-frequency data.

  In a tenth aspect of the noise determination apparatus according to the present invention, the frequency analysis unit further determines whether the high-frequency side data after the normalization is larger than a predetermined second threshold value. Analyzing the presence or absence of noise.

  An eleventh aspect of the noise determination device according to the present invention uses Bayer data as the image data.

  An aspect of the noise determination method according to the present invention is a noise determination method for determining presence / absence of noise in image data given in frame units, the step (a) for frequency-decomposing the image data, and the step (a). For the frequency components obtained in step 1, a predetermined number of data are sampled for low frequency components with relatively low frequencies and high frequency components with relatively high frequencies, and noise is determined based on the ratio of high frequency data to low frequency data. (B) which determines the presence or absence of.

  According to the noise determination device according to the present invention, since the presence or absence of noise is determined based on the ratio of the high frequency side data to the low frequency side data, an image captured under any circumstances is relatively Easily analyze the presence or absence of noise.

It is a block diagram which shows the structure of the noise removal apparatus which has the noise determination apparatus of embodiment which concerns on this invention. It is a block diagram which shows the structure of the noise determination apparatus of embodiment which concerns on this invention. It is a flowchart explaining the noise determination process in the noise determination apparatus of embodiment which concerns on this invention. It is a figure which shows the matrix of the DCT coefficient corresponding to an 8x8 Bayer data block. It is a block diagram which shows the modification of the noise determination apparatus of embodiment which concerns on this invention. It is a flowchart explaining the noise determination process in the modification of the noise determination apparatus of embodiment which concerns on this invention. It is a figure which shows the example of a setting of a sampling area. It is a figure which shows the example of a setting of a sampling area.

<Embodiment>
FIG. 1 is a block diagram showing a configuration of a noise removal apparatus 100 having a noise determination apparatus according to an embodiment of the present invention.

  As shown in FIG. 1, the noise removing device 100 receives image data and determines whether or not the image data contains noise, and a filter circuit 2 that removes noise contained in the image data. And a selection circuit 3 that selects filtered image data output from the filter circuit 2 based on the noise determination result in the noise determination device 1 or selects and outputs image data that has not been filtered. have.

  In FIG. 2, the structure of the noise determination apparatus 1 is shown with a block diagram. As shown in FIG. 2, Bayer data is input to the noise determination device 1 as image data, and the Bayer data is temporarily stored in the line memory group 11. As will be described later, since the noise determination apparatus 1 processes image data for one frame in a matrix of 8 × 8 pixels or 4 × 4 pixels, the line memory group 11 has a case of 8 × 8 pixels. Includes 7 line memories, and in the case of a 3 × 3 pixel unit, 3 line memories are included.

  The Bayer data stored in the line memory group 11 is given to the DCT (Discrete Cosine) converter 12 in units of 4 × 4 pixels or 8 × 8 pixels in a first-in first-out (FIFO) process, which is a Bayer of 4 lines or 8 lines. Continues until data is provided.

  The DCT transform unit 12 (frequency resolving unit) performs discrete cosine transform on the given Bayer data (hereinafter referred to as Bayer data block) to perform frequency decomposition, and outputs a transform coefficient for the obtained frequency component. .

  Then, the frequency analysis unit 13 analyzes the conversion coefficient output from the DCT conversion unit 12 to analyze whether or not there is noise in units of 4 × 4 pixels or 8 × 8 pixels. And the analysis result is stored in the analysis result storage unit 14 including a memory or the like.

  When the analysis result for the image data for one frame is stored in the analysis result storage unit 14, the analysis result for the image data for one frame is given to the single frame noise determination unit 15.

  The single frame noise determination unit 15 determines that the image data for one frame is noise based on the number of Bayer data blocks analyzed as having noise among the analysis results for the image data for one frame. It is determined whether or not the image data has a frame, that is, whether or not the frame has noise.

  Then, the determination result in the single frame noise determination unit 15 is given to the multiple frame noise determination unit 16, and the final noise determination is performed based on how many frames have noise among the predetermined number of multiple frames. Then, a control signal based on the determination result is given to the selection circuit 3, and the selection operation of the image data in the selection circuit 3 is controlled.

  The noise determination apparatus 1 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), a wired logic, and the like, reads a program stored in the ROM, and reads the program into the CPU. As a result, the noise determination device 1 is functionally realized.

<Noise judgment processing>
Next, the noise determination process in the noise determination apparatus 1 will be described using the flowchart shown in FIG. Note that the following description is processing in the frequency analysis unit 13 and the following shown in FIG. 2, and DCT processing is a well-known technique, and thus description thereof is omitted.

  First, in step S1, the frequency analysis unit 13 converts an AC (alternating current) component into a DC (direct current) in a matrix of conversion coefficients (DCT coefficients) output from the DCT conversion unit 12 (corresponding to an 8 × 8 Bayer data block). ) The AC component is normalized by dividing by the component.

  Here, FIG. 4 shows a matrix of DCT coefficients corresponding to an 8 × 8 Bayer data block. In FIG. 4, the top (upper left corner) of the matrix is a DC component, the horizontal direction of the matrix represents the vertical frequency component (V), and the vertical direction of the matrix represents the horizontal frequency component (H). An AC component closer to the DC component has a lower frequency, and an AC component closer to a corner that forms a diagonal with the DC component of the matrix has a higher frequency.

  In FIG. 4, a sampling region of a low frequency component (low frequency component) that is an AC component closer to the DC component is defined as a sampling region LSR (low frequency sampling region), and the AC component on the side that forms a diagonal with the DC component. A sampling region of a high frequency component (high frequency component) is shown as a sampling region HSR (high frequency sampling region).

  Next, in step S2, the frequency analysis unit 13 compares the total number of normalized DCT coefficients in the sampling region HSR with respect to the total number of DCT coefficients normalized in the sampling region LSR (Σ low region) illustrated in FIG. The ratio (Σhigh region / Σlow region) is calculated and compared with a predetermined threshold A. If the ratio exceeds the threshold A, the process proceeds to step S3. If the ratio is equal to or less than the threshold A, it is determined that the Bayer data block to be processed has no noise, and the determination result is analyzed. The result is stored in the result storage unit 14 (step S11). Thereafter, the process proceeds to step S5.

  In FIG. 4, the number of frequency components in the sampling region LSR and the number of frequency components in the sampling region HSR are set to about 30% of the whole, but this is a significant result if the region is made too wide. However, it can be set arbitrarily.

  In step S3, the total number of normalized DCT coefficients in the sampling region HSR (Σhigh region) is compared with a predetermined threshold B. This is because, for example, in an image with little change in which there are many DC components and few AC components, there is a small amount of high-frequency components. There are measures to avoid this. As a result of the comparison in step S3, if it is determined that there are more than a predetermined number of high frequency components, it is determined that the determination in step S2 is accurate.

  Therefore, if it is determined in step S3 that the total number of normalized DCT coefficients in the sampling region HSR exceeds the threshold B, it is determined that the Bayer data block to be processed has noise, The determination result is stored in the analysis result storage unit 14 (step S4).

  Next, in step S5, the frequency analysis unit 13 determines whether or not the noise determination has been completed for all the Bayer data blocks constituting one frame, and when the noise determination has been completed for all the Bayer data blocks. The process proceeds to step S6, but if there is an unprocessed Bayer data block, the processes after step S1 are repeated.

  In step S6, the number of Bayer data blocks determined to have noise for Bayer data in one frame based on the determination result stored in the analysis result storage unit 14 in the single frame noise determination unit 15 is determined in advance. It is determined whether or not the threshold value is greater than a predetermined threshold value C. If it is determined that the number of Bayer data blocks having noise is greater than a predetermined threshold C, it is determined that the frame has noise alone, and the determination result is stored in the analysis result storage unit 14. Store (step S7). On the other hand, if it is determined in step S6 that the number of Bayer data blocks having noise is equal to or less than a predetermined threshold C, it is determined that the frame does not have noise by itself, and the determination result is analyzed. Store in the storage unit 14 (step S12).

  In this case, for example, when there is noise in 30% of all Bayer data blocks in one frame, the threshold value C can be set to an arbitrary number such as determining that the frame has noise.

  Thereafter, in step S7, the multiple frame noise determination unit 16 checks the number of frames determined to have noise in a single frame, the number of consecutive frames, etc., and whether the multiple frames satisfy a predetermined determination condition. It is determined whether or not (step S8). If the determination condition is satisfied, it is determined that the moving image has noise (step S9), and a predetermined control signal is given to the selection circuit 3 so that the data from which the noise has been removed is output (step S9). Step S10). Thereafter, the power of the noise removing apparatus 100 is turned off, for example, and the processes in and after step S1 are repeated until no image data is input.

  On the other hand, if it is determined in step S8 that the predetermined determination condition is not satisfied, it is determined that the moving image has no noise (step S13), and a predetermined control signal is given to the selection circuit 3 (step S13). S10). The control signal is output both when it is determined that there is noise and when it is determined that there is no noise. For example, the control signal when there is noise is “0”, and there is no noise. By setting the control signal to “1”, the selection circuit 3 performs a selection operation corresponding to each.

  Here, the determination operation in step S8 will be described. Usually, a moving image is composed of 60 frames per second. Accordingly, in step S8, determination is made after the processing of steps S1 to S7 and S12 is completed for each of the 60 frames of image data. In this determination, as a determination condition, for example, when more than half of the 60 frames have noise, it is assumed that the moving image includes noise, and for the image data from the 61st frame onwards. The control signal for controlling the selection operation of the selection circuit 3 is output so that the filtered image data from which noise has been removed through the filter circuit 2 is output through the selection circuit 3.

  In other words, until 60 frames of image data necessary for the final determination are collected, image data that has not been subjected to filter processing is output through the selection circuit 3.

  Thereafter, the image data after the 61st frame becomes the image data after filtering, but the processing of steps S1 to S13 is repeated, and the number of frames that do not include noise increases, resulting in an imaging situation that does not satisfy the above-described determination condition. In this case, unfiltered image data is output through the selection circuit 3.

  The reason why such processing is performed is that there is a large difference in the captured image between when the filtering process is performed and when the filtering process is not performed, but when the filtering process is frequently turned on and off, the captured image flickers. This is to avoid this.

  Note that the determination condition is not limited to the number of frames determined to have noise as described above, and when a predetermined number of frames determined to have noise are continuous, for example, If 10 frames or more out of 60 frames are continuous, it is assumed that the moving image includes noise. For image data after the 61st frame, filtered image data from which noise has been removed through the filter circuit 2 is included. A control signal for controlling the selection operation of the selection circuit 3 may be output so as to be output through the selection circuit 3.

  Here, the processing of the above steps S1 to S3 is performed for each of the four colors Gb (green on the blue side), Gr (green on the red side), R (red), and B (blue) of the input Bayer data. By executing, the accuracy of determination can be increased. Note that since the green component includes luminance information, the processes of steps S1 to S3 may be executed for only Gb data or only Gr data.

<Modification of Noise Determination Processing Device>
Next, a modified example of the noise determination device 1 will be described. FIG. 5 is a block diagram of a noise determination apparatus 1 </ b> A that is a modification of the noise determination apparatus 1.

  As shown in FIG. 5, Bayer data is input as image data to the noise determination device 1 </ b> A, and the Bayer data is temporarily stored in the line memory group 11. As will be described later, since the noise determination apparatus 1 processes image data for one frame in units of 8 × 8 pixels or 4 × 4 pixels, the line memory group 11 is 7 in the case of 8 × 8 pixels. In the case of a 3 × 3 pixel unit, three line memories are included.

  The Bayer data stored in the line memory group 11 is given to the DCT (Discrete Cosine) converter 12 in units of 4 × 4 pixels or 8 × 8 pixels in a first-in first-out (FIFO) process, which is a Bayer of 4 lines or 8 lines. Continues until data is provided.

  The DCT transform unit 12 performs discrete cosine transform on the given 8 × 8 pixel Bayer data block and outputs transform coefficients.

  Then, the frequency analysis unit 13 performs frequency analysis on the conversion coefficient output from the DCT conversion unit 12 to analyze whether noise is present in units of 4 × 4 pixels or 8 × 8 pixels. A result is obtained, and the analysis result is stored in the analysis result storage unit 14.

  When the analysis result for one frame of image data is stored in the analysis result storage unit 14, the analysis result for the one frame of image data is given to the multiple frame noise determination unit 16, and among the predetermined number of multiple frames, the noise The final noise determination is performed based on the number of frames having “”, and a control signal based on the determination result is given to the selection circuit 3 to control the selection operation of the image data in the selection circuit 3.

  Next, the noise determination process in the noise determination apparatus 1A will be described using the flowchart shown in FIG.

  First, in step S21, the frequency analysis unit 13 converts an AC (alternating current) component into a DC (direct current) component in a matrix of conversion coefficients (DCT coefficients) output from the DCT conversion unit 12 (corresponding to an 8 × 8 Bayer data block). ) The AC component is normalized by dividing by the component.

  Next, in step S22, the frequency analysis unit 13 adds a DCT coefficient for each corresponding frequency component for each normalized matrix of DCT coefficients of all the Bayer data blocks constituting one frame. As a result, one matrix (data aggregation matrix) in which the DCT coefficients of all the matrices are aggregated is obtained.

  Next, in the data aggregation matrix, the frequency analysis unit 13 compares the total number of normalized DCT coefficients in the high frequency sampling region (Σ low frequency) with respect to the total number of DCT coefficients normalized in the low frequency sampling region (Σ low frequency). The ratio (Σhigh region / Σlow region) is calculated and compared with a predetermined threshold A. If the ratio exceeds the threshold A, the process proceeds to step S24. If the ratio is equal to or less than the threshold A, it is determined that the processing target frame does not have noise, and the determination result is stored as an analysis result. The data is stored in the unit 14 (step S29). Thereafter, the process proceeds to step S26.

  In step S14, the frequency analysis unit 13 compares the total number of the normalized DCT coefficients in the high frequency sampling region (Σ high frequency) with a predetermined threshold value B. This is because, for example, in an image with little change in which there are many DC components and few AC components, there is a small amount of high-frequency components, and it may be erroneously determined that there is noise only in the determination step of Step S23. There are measures to avoid this. As a result of the comparison in step S24, if it is determined that there are more than a predetermined number of high frequency components, it is determined that the determination in step S23 is accurate.

  Therefore, if it is determined in step S24 that the total number of normalized DCT coefficients in the high-frequency sampling region exceeds the threshold value B, it is determined that the processing target frame has noise, and the determination is made. The result is stored in the analysis result storage unit 14 (step S25).

  Then, in step S25, the number of frames stored as having noise, the number of consecutive frames, and the like are confirmed in the multiple frame noise determination unit 16, and whether or not the multiple frames satisfy a predetermined determination condition. Judgment is made (step S26). If the determination condition is satisfied, it is determined that the moving image has noise (step S27), and a predetermined control signal is given to the selection circuit 3 so that the data from which the noise has been removed is output (step S27). Step S28). After that, the process from step S21 onward is repeated until no image data is input, for example, when the power of the noise removing apparatus 100 is turned off.

  On the other hand, if it is determined in step S26 that the predetermined determination condition is not satisfied, it is determined that there is no noise in the moving image (step S30), and a predetermined control signal is given to the selection circuit 3 (step S30). S28). The control signal is output both when it is determined that there is noise and when it is determined that there is no noise. For example, the control signal when there is noise is “0”, and there is no noise. By setting the control signal to “1”, the selection circuit 3 performs a selection operation corresponding to each.

  Note that the determination operation in step S26 is the same as the determination operation in step S8 described above with reference to FIG.

  Here, the processes in steps S21 to S24 are performed for each of the four colors Gb (green on the blue side), Gr (green on the red side), R (red), and B (blue) of the input Bayer data. By executing, the accuracy of determination can be increased. Since the green component includes luminance information, the processes in steps S21 to S24 may be executed only for Gb data or only for Gr data.

  In addition, although the noise determination apparatus 1 and 1A mentioned above may be used independently, you may use both in parallel and may perform comprehensive noise determination based on the noise determination processing result each performed.

  That is, since the noise determination processing in the noise determination apparatus 1 determines noise in units of Bayer data blocks, accurate noise determination is possible even in a complex image where there is a biased region where noise determination is difficult. There is a feature that there is.

  On the other hand, the noise determination process in the noise determination apparatus 1A aggregates all areas of one frame and determines noise with one Bayer data block. Therefore, the number of processing steps is small, and noise determination can be performed relatively easily. Therefore, there is a feature that it is suitable for a case where noise exists uniformly in a simple image.

  Therefore, by using the noise determination devices 1 and 1A, it is possible to have both of the above two characteristics, and it is possible to perform more accurate noise determination by complementing the disadvantages.

<Sampling area setting>
As described above, the number of frequency components in the sampling region can be set arbitrarily, but can be set based on the average value of Gb and Gr data in the entire frame.

  That is, since the green component includes luminance information, the effects described below can be obtained by setting the sampling area according to the luminance (brightness) of the image.

  For example, when the luminance of the image is high, the S / N ratio increases, so the number of samples can be reduced, and both the sampling area LSR and the sampling area HSR are reduced as shown in FIG. By reducing the sampling area, the amount of data to be processed can be reduced and the processing speed can be increased.

  On the other hand, when the luminance of the image is low, the S / N ratio is small. Therefore, by increasing both the sampling area LSR and the sampling area HSR as shown in FIG. Can be suppressed.

  4, 7, and 8 show examples in which the sampling region LSR and the sampling region HSR are set to have the same number of frequency components, but either one is larger and the other is smaller. May be set.

<Effect>
As described above, in the embodiment according to the present invention, the frequency analysis unit 13 calculates the Σ high range / Σ low range and analyzes the presence / absence of noise, so the presence / absence of noise is relatively easily analyzed. it can. Further, since noise determination is performed using Bayer data as image data, the accuracy of noise detection can be improved as compared with the case where Bayer data is taken in and data after color conversion is performed by interpolation processing.

  When the above-described noise determination device is mounted on an image processing apparatus or the like together with a codec, the DCT unit and the quantization unit of the codec can be shared, so that the presence or absence of noise can be determined with a simpler circuit configuration. There is an effect.

<Other application examples>
In the embodiments according to the present invention described above, the configuration for performing noise determination using Bayer data as image data has been described. However, noise determination may be performed using YCbCr data or YUV data.

  In the above description, an example in which DCT is used for frequency decomposition has been described. However, noise determination may be performed using FFT (Fast Fourier Transform), wavelet transform, or the like.

  In the present invention, the embodiments can be appropriately modified and omitted within the scope of the invention.

12 Frequency Decomposition Unit 13 Frequency Analysis Unit 15 Single Frame Noise Determination Unit 16 Multiple Frame Noise Determination Unit

Claims (12)

A noise determination device for determining the presence or absence of noise in image data given in units of frames,
The image data is frequency-decomposed, and for the obtained frequency component, a predetermined number of data is sampled for each of a low frequency component having a relatively low frequency and a high frequency component having a relatively high frequency, and the high frequency side data for the low frequency side data is sampled A noise determination device that determines the presence or absence of noise based on the ratio. The noise determination device
A frequency resolving unit that divides the image data of one frame into a matrix of a predetermined number of pixels and performs frequency decomposition as a data block;
For the frequency components obtained by the frequency resolving unit, a predetermined number of frequency components are sampled on each of the low frequency component side and the high frequency component side of the data block to obtain the low frequency side data and the high frequency side data. The noise determination according to claim 1, further comprising: a frequency analysis unit that analyzes presence / absence of noise based on whether a ratio of the high frequency side data to the low frequency side data is greater than a predetermined first threshold value. apparatus. The noise determination device
A single frame noise determination unit that determines that the one frame has noise when the number of the data blocks analyzed as having noise in the frequency analysis unit is greater than a predetermined threshold; Item 3. The noise determination device according to Item 2. The noise determination device
The frame single noise determination unit includes a multiple frame noise determination unit that determines that a moving image has noise when a plurality of frames determined to have noise satisfy a predetermined determination condition. 3. The noise determination device according to 3. The noise determination device
A frequency resolving unit that divides the image data of one frame into a matrix of a predetermined number of pixels and performs frequency decomposition as a data block;
For the frequency components obtained by the frequency resolving unit, a predetermined number of frequency components are sampled on each of the low frequency component side and the high frequency component side of the data block to obtain the low frequency side data and the high frequency side data. Creating a data aggregation matrix in which data is added for each corresponding frequency component for all the data blocks constituting the one frame, and the high frequency side data is compared with the low frequency side data in the data aggregation matrix. The noise determination apparatus according to claim 1, further comprising: a frequency analysis unit that analyzes presence / absence of noise based on whether the ratio is greater than a predetermined first threshold value. The noise determination device
6. The multi-frame noise determination unit according to claim 5, further comprising: a plurality of frames determined to have noise in the frequency analysis unit when a predetermined determination condition is satisfied, wherein the frame is determined to have noise in the moving image. Noise judgment device. The determination condition is as follows:
The noise determination apparatus according to claim 4 or 6, which is defined by a ratio of frames having noise among a plurality of frames constituting a moving image. The determination condition is as follows:
The noise determination apparatus according to claim 4 or 6, which is defined by a continuous number of frames having noise among a plurality of frames constituting a moving image. The frequency resolving unit is
Perform frequency decomposition by discrete cosine transform,
The frequency analysis unit
The AC component is normalized by dividing the AC component of the transform coefficient obtained by the discrete cosine transform in the frequency resolution unit by the DC component, and the AC component after normalization is converted to the low frequency side data and the high frequency side. The noise determination device according to claim 2 or 5, wherein the data is data. The frequency analysis unit
The noise determination apparatus according to claim 9, further determining whether or not the high-frequency data after the normalization is greater than a predetermined second threshold, and analyzing the presence or absence of noise.   The noise determination apparatus according to claim 1, wherein Bayer data is used as the image data. A noise determination method for determining the presence or absence of noise in image data given in units of frames,
(A) frequency-decomposing the image data;
(B) For the frequency component obtained in the step (a), a predetermined number of data is sampled for each of the low frequency component having a relatively low frequency and the high frequency component having a relatively high frequency, and the high frequency side with respect to the low frequency side data is sampled. And a step of determining the presence or absence of noise based on the ratio of data.

Images