JP2008060722A - Image processor and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor and an image processing method which determine an image and are capable of processing the image on the basis of a determination result irrespective of a upconversion system by which the image has been upconverted. <P>SOLUTION: The image processor includes an image type determination unit 1200 and an image processing unit 1300 for processing an image on the basis of an image determination result of the image type determination unit 1200. The image type determination unit 1200 includes: a frequency band signal detecting unit 110 for detecting signals belonging to a plurality of frequency bands, from an image signal; an average value calculating unit for calculating average values of characteristic values corresponding to amplitudes for respective ones of the signals belonging to the plurality of frequency bands detected by the frequency band signal detecting unit 110; a relative value calculating unit 140 for calculating a relative value of one of the average values to another of the average values, among the average values calculated for respective ones of the plurality of frequency bands; an image determination unit 150 for determining the image based on the relative value calculated by the relative value calculating unit 140. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and an image processing method.

  In recent years, with the start of digital broadcasting, HD (High Definition) resolution image signals are transmitted from broadcast stations and the like. However, the image signals transmitted from broadcasting stations are not all HD resolution image signals, but SD (Standard Definition) resolution image signals used in conventional analog broadcasting are up-converted to HD resolution. In some cases, a pseudo HD resolution image signal is included.

  Known methods for up-converting an SD resolution image to an HD resolution image include a side panel method, a top and bottom crop method, and a stretch method.

  Generally, a pseudo HD resolution image obtained by up-converting an SD resolution image into an HD resolution image is inferior in terms of image quality as compared with a pure HD resolution image. It is necessary to enhance the image outline using (Image Enhancer).

  Under such circumstances, a technique has been developed for determining whether an image signal is an image obtained by up-converting an SD resolution image into an HD resolution image and performing image processing based on the determination result.

  As a technique for detecting a side panel to determine an image and performing image processing based on the determination result, for example, Patent Document 1 is cited. Further, as a technique for performing image processing by calculating average luminance in addition to detection of a side panel, for example, Patent Document 2 can be cited.

JP-A-2005-65195 JP-A-2005-26814

  However, the above-described technique for detecting an image by detecting a side panel requires that the signal of the side panel as an invalid area is only a signal equal to or less than a predetermined threshold. Therefore, the technique for determining the image by detecting the side panel and performing image processing based on the determination result is a signal of the side panel added to the image obtained by up-converting the SD resolution image to the HD resolution image. However, an image can be determined only when the signal is equal to or lower than a predetermined threshold, and image processing cannot be performed.

  In addition, the method of up-converting an SD resolution image to an HD resolution image is not only a side panel method but also a top-and-bottom crop method, a stretch method, etc., which do not include a side panel as an invalid area. There is. Therefore, the technique for determining the image by detecting the side panel described above and performing the image processing based on the determination result cannot determine the image when the side panel is not included in the image signal. This process cannot be performed.

  In addition to the above-described detection of the side panel, the technique for calculating the average luminance and performing the image processing is also based on the assumption that the side panel is detected. Cannot be processed.

  Therefore, the conventional technique for determining whether an image is an image obtained by up-converting an SD resolution image into an HD resolution image, and performing image processing based on the determination result, Image determination can be performed only in a limited manner, and image processing cannot be performed.

  Therefore, the present invention has been made in view of the above problems, and an object of the present invention is to determine an image regardless of what up-conversion method the image is up-converted, and to determine the result of the determination. It is an object of the present invention to provide a new and improved image processing apparatus and an image processing method capable of processing an image based on the above.

  In order to achieve the above object, according to a first aspect of the present invention, a frequency band signal detection unit for detecting signals in a plurality of frequency bands from an image signal, and a plurality of frequencies detected by the frequency band signal detection unit. An average value calculation unit that calculates an average value for a characteristic value corresponding to an amplitude for each signal in the band, and the other average for one average value among the average values calculated for each of the plurality of frequency bands An image type determining unit including a relative value calculating unit that calculates a relative value of values, an image determining unit that determines an image based on the relative value calculated by the relative value calculating unit, and an image of the image type determining unit An image processing apparatus including an image processing unit that processes an image based on the determination result is provided.

  The image processing apparatus includes an image type determination unit and an image processing unit. The image type determination unit includes a frequency band signal detection unit, an average value calculation unit, a relative value calculation unit, and an image determination unit, and determines an image.

  The frequency band signal detection unit detects signals in a plurality of frequency bands from the image signal, for example, by filtering the image signal for each pixel. Here, the signals in the plurality of frequency bands may be signals in a frequency band having a predetermined lower limit and an upper limit, or may be signals in a frequency band not lower than a predetermined lower limit without an upper limit.

  The average value calculation unit calculates an average value for the characteristic value corresponding to the amplitude for each of the signals in the plurality of frequency bands detected by the frequency band signal detection unit for each pixel. Here, the average value calculated by the average value calculation unit may be an arithmetic average of absolute values of characteristic values corresponding to the amplitude for each signal in a plurality of frequency bands.

  Further, the relative value calculation unit calculates a relative value of another average value with respect to one average value among the average values calculated for each of the plurality of frequency bands. Here, the lower limit of the frequency in the frequency band derived from the one average value is set lower than the lower limit of the frequency in the frequency band derived from the other average value.

  The image determination unit determines an image based on the relative value calculated by the relative value calculation unit. Here, the determination of the image may be uniquely determined by comparing a predetermined threshold value and the relative value to determine whether the image is an HD resolution image or an SD resolution image. In addition, the relative value itself can be used as an index of the probability as an HD resolution image.

  The image processing unit processes the image based on the image determination result of the image type determination unit. Here, the processing of the image in the image processing unit includes a process of enhancing the contour of the image based on whether or not the image type determination unit determines that the image is an image obtained by up-converting an image with SD resolution. The content of the processing can be changed as appropriate based on the relative value as an index of the certainty as an HD resolution image.

  The image type determination unit further includes a noise removing unit that removes a signal having the characteristic value less than a predetermined threshold for each of signals in a plurality of frequency bands detected by the frequency band signal detection unit, and the average value The calculation unit may calculate an average value for the characteristic value from which noise has been removed by the noise removal unit.

  The noise removing unit removes noise for each of a plurality of frequency band signals detected by the frequency band signal detecting unit for each pixel. Here, noise is removed by removing a signal having the above characteristic value less than a predetermined threshold. The average value calculation unit calculates an average value for the characteristic value from which noise has been removed by the noise removal unit. By providing the noise removing unit, noise, that is, an abnormal value can be removed, so that the image type determining unit can perform more accurate image determination.

  The relative value calculation unit is based on the average value calculated in a predetermined frequency band determined in advance and the average value calculated in a frequency band higher than the predetermined frequency band determined in advance. Thus, the relative value may be calculated.

  The relative value calculation unit calculates a relative value of the average value calculated in a frequency band higher than the predetermined frequency band with respect to the average value calculated in a predetermined frequency band. To do. Therefore, one relative value is output from the relative value calculation unit. The image type determination unit determines an image using the one relative value.

  The lower limit frequency of the predetermined frequency band is lower than the theoretically effective upper limit frequency according to the upconversion method with respect to the Nyquist frequency of the image upconverted from the standard resolution image, and The lower limit frequency of the high frequency band may be higher than the upper limit of the theoretically effective frequency corresponding to the up-conversion method with respect to the Nyquist frequency of the image up-converted from the standard resolution image.

  A general image has a frequency amplitude characteristic in which the amplitude is large in a low frequency band and the amplitude is attenuated as the frequency is increased. Therefore, a relative value more suitable for image determination can be calculated by using the upper limit of the theoretically effective frequency for the Nyquist frequency of the image up-converted from the standard resolution (SD resolution) image as a reference. It becomes possible. Here, the Nyquist frequency of the image up-converted from the standard resolution (SD resolution) image is uniquely determined according to the sampling frequency (Nyquist frequency = sampling frequency × (1/2)), and is theoretically effective for the Nyquist frequency. The upper limit of the correct frequency is uniquely determined according to the up-conversion method.

  Further, the relative value calculation unit is configured to calculate the average value calculated in a predetermined frequency band, and the average value calculated in a plurality of frequency bands higher than the predetermined frequency band. A plurality of the relative values may be calculated based on the above.

  The relative value calculation unit is a relative value of the average value calculated in a plurality of frequency bands higher than the predetermined frequency band with respect to the average value calculated in a predetermined frequency band. A value is calculated for each of the plurality of frequency bands. Therefore, a plurality of relative values are output from the relative value calculation unit. The image type determination unit determines an image using the plurality of relative values. Here, the image type determination unit may determine the image based on one of the plurality of relative values, or may determine the image based on the plurality of relative values. .

  Further, the relative value calculation unit is calculated in a higher frequency band than the average value calculated in a plurality of predetermined frequency bands and the predetermined frequency bands. A plurality of the relative values may be calculated based on the average value.

  The relative value calculation unit predetermines a plurality of predetermined frequency bands, and more than each of the predetermined plurality of predetermined frequency bands with respect to the average value calculated in the predetermined plurality of predetermined frequency bands, A relative value of the average value calculated in the high frequency band is calculated for each of the plurality of predetermined frequency bands. Therefore, the relative value calculation unit outputs a plurality of relative values having different frequency bands serving as a reference for calculating the relative value. The image type determination unit determines an image using the plurality of relative values. Here, the image type determination unit may determine the image based on one of the plurality of relative values, or may determine the image based on the plurality of relative values. .

  Further, the average value calculation unit calculates the average value in each frequency band of the plurality of frequency bands for each local region in the entire image, and the image determination unit calculates the local value based on the relative value. The image may be determined for each region.

  The average value calculation unit calculates the average value in each frequency band of the plurality of frequency bands for each local region in the entire image. The image determination unit determines an image based on a relative value calculated using the average value calculated for each local region. Therefore, the image type determination unit performs the determination for each local region in the entire image even if the image to be determined includes an HD resolution image portion and an SD resolution image portion upconverted. Therefore, the image can be accurately determined.

  The image type determination unit may further include a time smoothing unit that time-smooths the relative value, and the image determination unit may determine an image based on the time-smoothed relative value.

  The image type determination unit further includes a time smoothing unit, and time smoothes the relative value. Since the image determination unit determines an image based on the time-smoothed relative value, for example, when the relative value largely fluctuates due to some factor even though the scene of the image does not change Thus, the fluctuation of the relative value can be suppressed to the minimum, and the image determination can be performed while suppressing the variation of the image determination result in the same scene.

  In addition, the image processing apparatus further includes a scene change detection unit that detects a change in the scene of the image represented by the image signal, and the time smoothing unit detects the time point when the scene change is detected in the scene change detection unit. To time smoothing of the relative value.

  The image processing apparatus further includes a scene change detection unit. The scene change detection unit detects a change in the scene of the image represented by the image signal, and the time smoothing unit detects the relative change from the point in time when the scene change is detected in the scene change detection unit. Performs time smoothing of values. Time smoothing of the relative value optimal for each scene is performed by newly performing the time smoothing of the relative value from the time when the change of the scene is detected, that is, from the time when the relative value of the image changes. Therefore, it is possible to accurately determine an image without being affected by a scene change.

  In order to achieve the above object, according to a second aspect of the present invention, a step of detecting a signal of a plurality of frequency bands from an image signal, and an amplitude for each of the detected signals of the plurality of frequency bands A step of calculating an average value for a characteristic value corresponding to the step, a step of calculating a relative value of another average value with respect to one of the average values among the average values calculated for the plurality of frequency bands, and There is provided an image processing method including a step of determining an image based on the calculated relative value, and a step of processing the image based on a determination result of the image.

  By using this method, it is possible to determine whether the image is an HD resolution image or a pseudo HD resolution image obtained by up-converting an SD resolution image, and the image is processed based on the determination result. be able to.

  According to the present invention, it is possible to determine an image regardless of which up-conversion method is used for the image, and to process the image based on the determination result.

  Exemplary embodiments of the present invention will be described below in detail with reference to the accompanying drawings. In the present specification and drawings, components having substantially the same functional configuration are denoted by the same reference numerals, and redundant description is omitted.

(Principle of image judgment)
First, the principle of image determination in the present invention will be described.

  FIG. 1 is an explanatory diagram showing an HD resolution image and an image obtained by up-converting an SD resolution image. Here, FIG. 1A shows an HD resolution image. 1B shows an SD resolution image, and FIG. 1C shows an image obtained by up-converting the SD resolution image (hereinafter referred to as “pseudo HD resolution image”). Yes.

  Also, as shown in FIG. 1C, a side panel as an invalid area is not always added to a pseudo HD resolution image. Therefore, the conventional image determination technique cannot determine a pseudo HD resolution image that does not include a side panel as shown in FIG. Therefore, the present invention uses a difference in frequency amplitude characteristics between an HD resolution image and a pseudo HD resolution image in order to discriminate between an HD resolution image and a pseudo HD resolution image.

  FIG. 2 is an explanatory diagram for explaining the difference in frequency amplitude characteristics between the HD resolution image shown in FIG. 1A and the pseudo HD resolution image shown in FIG. FIG. 2 shows the relationship between frequency and amplitude in an image, with the normalized angular frequency on the horizontal axis and the power spectrum of the amplitude on the vertical axis.

  Referring to FIG. 2, the HD resolution image shown in FIG. 1 (a) and the pseudo HD resolution image shown in FIG. 1 (c) have a low frequency band in FIG. ”)), It can be seen that the amplitude of the signal is attenuated as the frequency is increased and the signal is increased. Here, the frequency amplitude characteristic in which the amplitude is large in the low frequency band as shown in FIG. 2 and the amplitude is attenuated as the frequency is increased is established for a general image such as a natural image.

  Here, when the HD resolution image signal and the pseudo HD resolution image signal are compared with each other based on FIG. 2, the signal is similarly distributed in the low frequency band. It can be seen that the attenuation rate of the amplitude is different. That is, in a band higher in frequency than the low frequency band in FIG. 2 (hereinafter referred to as “middle frequency band”), the signal present in the area A in FIG. There are very few HD resolution image signals. Therefore, the amplitude attenuation rate with respect to the low frequency band is larger in the pseudo HD resolution image than in the HD resolution image.

  Further, in a region where the frequency is higher than the middle frequency band in FIG. 2 (hereinafter referred to as “high frequency band”), the pseudo HD resolution image signal is hardly distributed, and the HD resolution image is not distributed. The difference is that the signal is distributed. Note that the frequency amplitude characteristics shown in FIG. 2 described above are not limited to the images shown in FIGS. 1A and 1C, and almost all the images have the same tendency.

  Therefore, the present invention pays attention to the difference in attenuation rate with respect to the low frequency band between the HD resolution image and the pseudo HD resolution image, and the index (hereinafter referred to as “HD degree”) shown in Equation 1 is the image. Used to determine

  HD degree = (| average value of characteristic value corresponding to amplitude in frequency band 2) / (| average value of characteristic value corresponding to amplitude in frequency band 1) (Equation 1)

  Here, the lower limit frequency of frequency band 1 in Equation 1 is set to a frequency lower than the lower limit frequency of frequency band 2. Therefore, from the frequency amplitude characteristic shown in FIG. 2, the HD degree takes a value of 0 ≦ HD degree <1. Here, the HD degree becomes 0 (zero) when the characteristic value corresponding to the amplitude (hereinafter referred to as “amplitude characteristic value”) does not exist in the frequency band 2.

  Here, the amplitude characteristic value may be the amplitude itself or an amplitude power spectrum. Needless to say, the amplitude characteristic value is not limited to the above. Moreover, although the average value shown in Numerical formula 1 can be calculated | required by an arithmetic mean, it is not restricted to the calculation method which concerns, It is good also as a geometric mean, It can also be set as the weighted average which carried out predetermined weighting.

  In Equation 1, the HD degree is calculated based on an average value with respect to the absolute value of the amplitude characteristic value in order to obtain the HD degree more accurately. However, the present invention is not limited to such a calculation method, and the amplitude before the absolute value is calculated. Needless to say, the HD degree can be calculated based on the characteristic value.

  Since the HD degree is a relative value between the average value of the amplitude characteristic values in the frequency band 1 and the average value of the amplitude characteristic values in the frequency band 2, the HD resolution image and the pseudo HD resolution shown in FIG. The difference in attenuation rate from the image can be expressed. Therefore, the higher the degree of HD of the image to be determined is closer to 0 (zero), the higher the probability that the image to be determined is a pseudo HD resolution image. It can be said that as the value closer to 1 is taken, the image to be determined is an HD resolution image.

(First embodiment)
Next, an image processing apparatus according to the first embodiment of the present invention will be described. FIG. 3 is an explanatory diagram showing the image processing apparatus 1000 according to the first embodiment of the present invention.

  Referring to FIG. 3, the image processing apparatus 1000 according to the first embodiment of the present invention includes an image signal receiving unit 1100, an image type determination unit 1200, an image processing unit 1300, and a display unit 1400. The image signal receiving unit 1100 receives an image signal transmitted from a broadcasting station or the like, or an image signal of an image held in a storage unit (not shown). The image type determination unit 1200 determines whether the image represented by the image signal is an HD resolution image or a pseudo HD resolution image. The image determination in the image type determination unit 1200 will be described later.

  The image processing unit 1300 performs image processing based on the image determination result in the image type determination unit 1200. Here, for example, when the image type determination unit 1200 determines that the image is a pseudo HD resolution image, the image processing unit 1300 performs image processing with predetermined parameters, and the image is an HD resolution image. When it is determined that the image is processed, the image processing can be performed based on the binary determination result, for example, the image processing is not performed.

  In addition, the image processing unit 1300 can select a parameter for processing based on the value of the HD degree calculated by the image type determination unit 1200 and change the processing as appropriate. Here, the selection of the parameter can be realized by the image processing unit 1300 including a parameter selection unit (not shown) for selecting a parameter. Alternatively, the image processing apparatus 1000 may include a parameter selection unit (not shown) that is separate from the image processing unit 1300.

  Note that the image processing method performed by the image processing unit 1300 is not limited to the above. For example, a plurality of HD degree values are calculated from the image type determination unit 1200, and the plurality of HD degree values are used. It goes without saying that the above processing can also be performed.

  The image processing unit 1300 also performs processing for enhancing the contour of the image (hereinafter referred to as “enhancement processing”), contrast processing, and noise reduction for reducing noise based on the image determination result in the image type determination unit 1200. And so on. Needless to say, the image processing unit 1300 can perform various processes without being limited to the above.

  The display unit 1400 displays the image processed by the image processing unit 1300. Here, the image displayed on the display unit 1400 is always an HD resolution image or a pseudo HD resolution image subjected to image processing. Therefore, a high-quality image is displayed on the display unit 1400.

(Image processing method)
Next, an image processing method according to an embodiment of the present invention will be described. FIG. 4 is a flowchart showing an image processing method according to the embodiment of the present invention.

  First, signals in a plurality of frequency bands are detected from the input image signal (S100). Here, the detection of the signal is performed for each pixel.

  For each signal in a plurality of frequency bands detected for each pixel in step S100, an average value for the amplitude characteristic value is calculated (S102). Here, the amplitude characteristic value may be the amplitude value itself or a power spectrum. The calculation method of the average value is not limited to the arithmetic average, and various calculation methods such as a geometric average or a weighted average can be used.

  Of the average values calculated for each signal in a plurality of frequency bands in step S102, a relative value of another average value with respect to one average value is calculated (S104). Here, the relative value can be obtained by dividing the other average value by the one average value, or from the logarithm of the other average value, You may subtract the value which took the logarithm.

  Further, the lower limit frequency of the frequency band derived from the one average value is set lower than the lower limit frequency of the frequency band derived from the other average value. Therefore, the calculated relative value can represent the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image shown in FIG.

  An image is determined based on the relative value calculated in step S104 (S106). Here, as the relative value takes a value close to 0 (zero), the probability of being a pseudo HD resolution image is very high, and as the value close to 1 is taken, an HD resolution image becomes higher. The probability of being is very high. Therefore, based on a predetermined threshold, if the HD degree is equal to or greater than the threshold, it is determined that the image is an HD resolution image. If the HD degree is less than the threshold, the image is a pseudo HD resolution image. It can be determined that

  Note that the image determination in step S106 is not limited to the above, and it goes without saying that the HD degree value itself can be regarded as a probability value that is an HD resolution image, and the probability that the image is an HD resolution image can be determined. Yes.

  Based on the image determination result in step S106, image processing is performed (S108). Here, in step S108, enhancement processing, color adjustment, noise reduction, and the like can be performed. Note that the image processing in step S108 is not limited to the above, and it is needless to say that various processing can be performed.

  As shown in steps S100 to S108, the image processing method according to the embodiment of the present invention uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. The image is determined based on the relative value of the average value of the amplitude characteristic value in the higher frequency band with respect to the average value of the amplitude characteristic value in the predetermined predetermined frequency band, and the image is determined based on the determination result. Perform the process.

  Therefore, in the image processing method according to the embodiment of the present invention, it is not necessary to add a side panel at the time of up-conversion as in the conventional image processing method, and it depends on what type of up-conversion the image is up-converted. First, the image can be determined, and the image can be processed based on the determination result.

  As described above, the image processing apparatus according to the present invention uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image to determine whether the image has been upconverted. Therefore, since the image can be accurately determined, it is possible to perform a process suitable for the image. Then, next, the image type determination part which concerns on this invention using the principle of the image determination which concerns on this invention mentioned above is demonstrated.

[First Example of Image Type Determination Unit]
FIG. 5 is an explanatory diagram illustrating the image type determination unit 100 according to the first embodiment of the image type determination unit.

  Referring to FIG. 5, the image type determination unit 100 according to the first embodiment of the image type determination unit includes a frequency band signal detection unit 110, a noise removal unit 120, an average value calculation unit 130, and a relative value calculation unit. 140 and an image determination unit 150.

  The image type determination unit 100 receives an image signal received by the image signal receiving unit 1100, that is, an image signal transmitted from a broadcasting station or the like, or an image signal of an image held in a storage unit.

  The frequency band signal detection unit 110 includes first frequency band signal detection means 112 and second frequency band signal detection means 114. The first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 perform band-by-pixel processing, pass only an image signal in a specific frequency band, and attenuate image signals in other bands. A filter (Band-Pass Filter; hereinafter referred to as “BPF”) can be used.

  The first frequency band signal detecting means 112 passes a frequency band lower than the upper limit of the theoretically effective frequency for the Nyquist frequency in the image up-converted from the SD resolution image, and the DC (Direct Current) component is passed. It is composed of a low-pass BPF that does not pass. The second frequency band signal detection means 114 is configured by a BPF in the middle band that passes the vicinity of the upper limit of the theoretically effective frequency for the Nyquist frequency in the image up-converted from the SD resolution image. Here, the low frequency BPF constituting the first frequency band signal detecting means 112 and the middle frequency BPF constituting the second frequency band signal detecting means 114 may partially overlap with each other. The upper limit of the theoretically effective frequency for the Nyquist frequency in an image up-converted from an SD resolution image varies depending on the up-conversion method. That is, as shown in FIG. 2, when the normalized Nyquist frequency in the HD resolution and the pseudo HD resolution is a (a = π), the image of the up-converted image in the top & bottom crop method and the stretch method is used. The upper limit b of the theoretically effective frequency for the Nyquist frequency is b = (3/8) × a, and the upper limit c of the theoretically effective frequency for the Nyquist frequency of the image up-converted by the side panel method is c = (4/9) × a. Therefore, the first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 set the first frequency band signal detection unit 112, for example, with reference to the theoretically effective upper limit b for the Nyquist frequency. The second frequency band signal detection means 114 can be set based on the theoretically effective upper limit c for the Nyquist frequency. The first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 are not limited to the above, and both the first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 are Nyquist. The upper limit b of the theoretically effective frequency for the frequency may be set as a reference.

  Further, the BPF constituting the first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 needs to reduce the side lobes in order to increase the detection capability. Therefore, the BPF according to the first embodiment sets the filter coefficient value using a function obtained by multiplying the Sinc function by the Hamming window function, and sets the number of taps so as to correspond to the filter coefficient value. Set and configure BPF.

  By adopting the above configuration, the BPF according to the first embodiment of the image type determination unit reduces the side lobes and uses the short wave as the main lobe, rather than using the discrete Fourier transform. You can get closer. Needless to say, the window function used to set the filter coefficient value is not limited to the Hamming window function, and a Hanning window function or a Gaussian window function can be used. Further, the BPF according to the first embodiment of the image type determination unit is not limited to the above-described configuration, and may be configured with a digital filter.

  The first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 are not limited to the BPF, and pass only an image signal having a frequency equal to or higher than the cutoff frequency, and attenuate an image signal having a frequency lower than the cutoff frequency. It can also be configured with a high-pass filter. Further, the first frequency band signal detection means 112 and the second frequency band signal detection means 114 pass only an image signal having a frequency equal to or lower than the cutoff frequency, and a low-pass filter that attenuates an image signal having a frequency higher than the cutoff frequency ( (Low-Pass Filter). However, when a low-pass filter is used, it is affected by the average luminance corresponding to a signal having a frequency of 0 (zero), and the HD luminance image and the pseudo HD resolution image cannot be accurately determined by the average luminance. There is.

  Therefore, the frequency band signal detection unit 110 is preferably a filter that does not allow an image signal having a frequency of 0 (zero) to pass therethrough, and is not limited to a BPF, and can be configured with an arbitrary filter.

  The noise removing unit 120 includes first noise removing means 122 and second noise removing means 124. The first noise removing unit 122 removes noise of the image signal filtered by the first frequency band signal detecting unit 112 for each pixel, and outputs an amplitude characteristic value for each pixel. The second noise removing unit 124 removes noise from the image signal filtered by the second frequency band signal detecting unit 114 for each pixel.

  Here, the noise removal in the noise removal unit 120 can be performed by presetting a threshold value of an amplitude characteristic value regarded as noise for each frequency band that the frequency band signal detection unit 110 passes. Alternatively, the image type determination unit 100 is provided with a noise measurement unit (not shown) that detects noise for each frequency band that the frequency band signal detection unit 110 passes, and based on the noise measurement result measured by the noise measurement unit, Noise may be removed. Needless to say, the noise removing means in the noise removing unit 120 is not limited to the above.

  The average value calculation unit 130 includes first average value calculation means 132 and second average value calculation means 134. The amplitude characteristic value for each pixel from which noise has been removed by the first noise removing unit 122 is input to the first average value calculating unit 132, and an average value with respect to the absolute value of the amplitude characteristic value is calculated. Further, the amplitude characteristic value for each pixel from which the noise has been removed by the second noise removing unit 124 is input to the second average value calculating unit 134, and an average value with respect to the absolute value of the amplitude characteristic value is calculated.

  Here, the average value in the first average value calculating means 132 and the second average value calculating means 134 can be obtained by an arithmetic average. Needless to say, the average value calculation method in the average value calculation unit 130 is not limited to the arithmetic average, and can be calculated by various methods such as a geometric average and a weighted average.

  The relative value calculation unit 140 outputs the average value relative to the absolute value of the amplitude characteristic value output from the average value calculation unit 130, that is, the first average value calculation unit 132 and the second average value calculation unit 134, respectively. Calculate the value. Here, the relative value calculated by the relative value calculation unit 140 corresponds to the HD degree shown in Equation 1. Therefore, the calculation of the relative value in the relative value calculation unit 140 can be obtained by the following formula 2 according to the formula 1.

HD degree = relative value calculated by the relative value calculation unit 140 = (average value calculated by the second average calculation unit 134) / (average value calculated by the first average calculation unit 132) (Equation 2)

  Needless to say, the relative value calculated by the relative value calculation unit 140, that is, the HD degree calculation method is not limited to Formula 2, and may be expressed using, for example, a logarithm.

  The image determination unit 150 determines an image using the relative value calculated by the relative value calculation unit 140, that is, the HD degree. Here, as described above, the higher the degree of HD calculated by the relative value calculation unit 140 is close to 0 (zero), the higher the probability that the image to be determined is a pseudo HD resolution image. Become. Further, the higher the degree of HD calculated by the relative value calculation unit 140 is closer to 1, the higher the probability that the image to be determined is an image of HD resolution. Therefore, the image determination unit 150 determines that the image is an HD resolution image if the HD degree is equal to or greater than the threshold based on a predetermined threshold, and if the HD degree is less than the threshold, the image Can be determined as a pseudo HD resolution image.

  Note that image determination in the image determination unit 150 is not limited to the above, and it is possible to determine the probability of being an HD resolution image by regarding the HD degree value itself as a probability value that is an HD resolution image. Needless to say.

  Further, the image determination unit 150 outputs an image determination result. Here, the determination result output from the image determination unit 150 can be, for example, a digital signal indicating an HD resolution image and a pseudo HD resolution image. Needless to say, the determination result output from the image determination unit 150 is not limited to the above, and can be the value of the HD degree itself.

  As described above, the image type determination unit 100 according to the first embodiment of the image type determination unit uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Then, the image is determined based on the relative value of the average value of the amplitude characteristic values in the middle frequency band with respect to the average value of the amplitude characteristic values in the low frequency band.

  Therefore, the image type determining unit 100 according to the first embodiment of the image type determining unit adds a side panel when an SD resolution image is upconverted, as in the conventional image type determining unit. The determination of the image can be performed regardless of the up-conversion method by which the image is up-converted.

[Program for image determination]
The program for causing the image type determination unit 100 according to the first embodiment of the image type determination unit described above to function as a computer performs image determination regardless of the up-conversion method of the image. be able to.

[Second Example of Image Type Determination Unit]
In the image type determination unit 100 according to the first example of the image type determination unit, the configuration for calculating one relative value as the HD degree has been described. Next, a method for calculating a plurality of relative values as the HD degree will be described. FIG. 6 is an explanatory diagram illustrating an image type determination unit 200 according to the second embodiment of the image type determination unit.

  Referring to FIG. 6, the image type determination unit 200 according to the second embodiment of the image type determination unit includes a frequency band signal detection unit 210, a noise removal unit 220, an average value calculation unit 230, and a relative value calculation unit. 240 and an image determination unit 250.

  The basic functions and effects of the respective units of the image type determining unit 200 are the same as those of the image type determining unit 100 according to the first embodiment of the image type determining unit, but the image type determining unit 100 and the image type determining unit are the same. The difference from 200 is that the frequency band signal detection unit 210 includes first frequency band detection means 212 to nth frequency band detection means 218 (where n is an integer of 3 or more), that is, three. The above frequency band detection means is provided. Here, the frequency band detection means included in the frequency band signal detection unit 210 can be configured by BPF as with the image type determination unit 100, but the first frequency band detection means 212 and the second frequency band detection means 214. , Third frequency band detecting means,..., Nth frequency band detecting means 218, and configured such that the lower limit of the frequency that the BPF passes is also increased as n increases.

  Similarly, the difference between the image type determination unit 100 and the image type determination unit 200 is that the noise removal unit 220 includes a first noise removal unit 222 to an nth noise removal unit 228, and the average value calculation unit 230 has a first average value. The calculation means 232 to the n-th average value calculation means 238 are provided, and the relative value calculation unit 240 includes the first relative value calculation means 242 to the n-1th relative value calculation means 244. The image determination unit 250 receives a plurality of relative values.

  Here, the first relative value calculation unit 242 of the relative value calculation unit 240 calculates the relative value of the average value calculated by the second average value calculation unit 234 with respect to the average value calculated by the first average value calculation unit 232. Further, the (n−1) -th relative value calculating unit 244 calculates a relative value of the average value calculated by the n-th average value calculating unit 238 with respect to the average value calculated by the first average value calculating unit 232. Therefore, the relative value calculation unit 240 calculates a plurality of relative values based on the average value calculated by the first average value calculation unit 232.

  The image determination unit 250 determines an image using the plurality of relative values calculated by the relative value calculation unit 240. Here, the image determination unit 250 can determine an image based on one of the plurality of relative values. For example, the image determination unit 250 calculates a relative value calculated using a frequency band having the largest lower limit frequency among a plurality of relative values having a value larger than 0 (zero), that is, a relative value having the largest n. The image can be determined based on whether the relative value calculated by the means is greater than or less than a predetermined threshold. Needless to say, the determination of an image based on one relative value is not limited to the above, and any relative value calculated may be used.

  Here, as shown in FIG. 2, the higher the frequency, the smaller the difference between the average values of the amplitudes included in the HD resolution image and the pseudo HD resolution image. Accordingly, a plurality of threshold values used for image determination are provided in advance for each frequency band, and image determination is performed by appropriately changing the threshold value for each frequency band used for image determination.

  The image determination unit 250 can also determine an image based on a plurality of calculated relative values. For example, the image determination unit 250 performs image determination using the above-described threshold values for all of the plurality of calculated relative values, and the image determination unit 250 determines that the image determination result is larger. Good. Needless to say, image determination based on a plurality of relative values is not limited to the above, and an arbitrary number of calculated relative values may be used.

  The image determination in the image determination unit 250 is performed by, for example, calculating a difference value from a threshold used for image determination for each of a plurality of relative values calculated by the relative value calculation unit 240, and calculating the plurality of calculated differences. Calculate the mean and standard deviation of the values. Here, if the standard deviation does not become a value within a predetermined range, a value with a large difference value is excluded as an abnormal value, and the average value and standard deviation of the difference value are calculated again. To do. Then, the above-described operation is repeated until the standard deviation becomes a value within a predetermined range. When the standard deviation finally becomes a value within a predetermined range, if the difference average value is a positive value, the image is determined to be an HD resolution image, and the difference average If the value is a negative value, it can be determined that the image is a pseudo HD resolution image. It goes without saying that the image determination in the image determination unit 250 is not limited to the above.

  The image determination unit 250 outputs an image determination result. The determination result output from the image determination unit 250 is, for example, a digital signal indicating an HD resolution image and a pseudo HD resolution image, as in the image determination unit 150 according to the first embodiment of the image type determination unit. can do. The image determination unit 250 can also output a plurality of HD degrees having a value greater than 0 (zero).

  In addition to the digital signal, the image determination unit 250 selectively outputs a plurality of relative values having a value greater than 0 (zero), that is, one HD degree value among a plurality of HD degrees. May be. Here, when the determination result output from the image determination unit 250 is the value of the HD degree, for example, among the plurality of HD degrees having a value larger than 0 (zero), the relative value calculation unit having the largest n. The value of the HD degree calculated by is output. As shown in FIG. 2, a general image has a frequency amplitude characteristic in which the amplitude decreases as the frequency increases. Therefore, the image processing unit 1300 performs image processing such as enhancement processing on the highest frequency band in which the amplitude exists, thereby effectively improving the image quality. It goes without saying that the HD degree value selected by the image determination unit 250 can be the HD degree value in an arbitrary frequency band.

  As described above, the image type determination unit 200 according to the second embodiment of the image type determination unit uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Then, a plurality of relative values of the average value of the amplitude characteristic value in the higher frequency band with respect to the average value of the amplitude characteristic value in the low frequency band are calculated, and the image is determined using the plurality of relative values. .

  Therefore, the image type determination unit 200 according to the second embodiment of the image type determination unit adds a side panel when an SD resolution image is up-converted as in the conventional image type determination unit. The determination of the image can be performed regardless of the up-conversion method by which the image is up-converted.

  Further, since the image type determination unit 200 according to the second embodiment of the image type determination unit can perform image determination using a plurality of relative values, for example, an image having no signal in a specific frequency band. Even so, it is possible to perform image determination based on signals in other frequency bands.

[Third Example of Image Type Determination Unit]
Next, an image type determination unit according to a third embodiment of the image type determination unit will be described. FIG. 7 is an explanatory diagram illustrating an image type determination unit 300 according to the third embodiment of the image type determination unit.

  Referring to FIG. 7, the image type determination unit 300 according to the third example of the image type determination unit has the same configuration as the image type determination unit 200 according to the second example of the image type determination unit illustrated in FIG. 6. Compared with the image type determination unit 200, the reference value for calculating the relative value in the first relative value calculation unit 342 to the n-th relative value calculation unit 344 of the relative value calculation unit 340 is different.

  Specifically, the image type determination unit 300 uses the relative value of the average value calculated by the second average value calculation unit 234 with respect to the average value calculated by the first average value calculation unit 232 by the first relative value calculation unit 342. And the n-1th relative value calculation means 344 calculates the relative value of the average value calculated by the nth average value calculation means 238 with respect to the average value calculated by the n-1th average value calculation means 236. To do. Therefore, in the image type determination unit 200 according to the second embodiment of the image type determination unit, all relative values are calculated based on the average value calculated by the first average value calculation unit 232, whereas the image type determination unit 200 The image type determination unit 300 according to the third example of the type determination unit has a difference in that when calculating a relative value, a frequency band for deriving an average value serving as a reference for calculating the relative value is changed.

  The image type determination unit 300 according to the third embodiment of the image type determination unit calculates the relative value as described above, so that the frequency amplitude characteristic of the image does not show the characteristic as shown in FIG. Even when a special image that does not have an amplitude characteristic value at a low frequency is determined, for example, the image can be determined based on the amplitude characteristic value in the middle frequency band.

  The image type determination unit 300 is the same as the image type determination unit 200 according to the second example of the image type determination unit in the image determination method using the relative value and the output method of the determination result.

  As described above, the image type determination unit 300 according to the third embodiment of the image type determination unit uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Then, a plurality of reference frequency bands are provided, and the relative value of the average value of the amplitude characteristic values in the higher frequency band is calculated with respect to the average value of the amplitude characteristic values in the reference frequency band. Then, the image is determined using the plurality of calculated relative values.

  Therefore, the image type determination unit 300 according to the third embodiment of the image type determination unit adds a side panel when an SD resolution image is upconverted, as in the conventional image type determination unit. The determination of the image can be performed regardless of the up-conversion method by which the image is up-converted.

  Further, the image type determination unit 300 according to the third embodiment of the image type determination unit is limited to an image even if the image to be determined is a special image having no frequency amplitude characteristics as shown in FIG. The determination of the image can be performed without being performed.

[Fourth Example of Image Type Determination Unit]
In the first to third embodiments of the image type determination unit according to the present invention described above, an image type determination for determining whether the entire image to be determined is an HD resolution image or a pseudo HD resolution image. I explained the part. However, there are cases where an image signal transmitted from a broadcast station or the like contains both HD resolution images and pseudo HD resolution images. Therefore, an embodiment in which an image can be determined when an HD resolution image and a pseudo HD resolution image are mixed will be described.

  FIG. 8 is an explanatory diagram showing an image in which an HD resolution image and a pseudo HD resolution image are mixed. FIG. 8A is an explanatory diagram showing an image in which an HD resolution image and a pseudo HD resolution image are mixed, and FIG. 8B divides the image of FIG. 8A into local regions. FIG.

  As shown in FIG. 8A, for example, in a news program, an HD resolution image and a pseudo HD resolution image may be mixed. In the case of an image as shown in FIG. 8A, it is assumed that an accurate determination result cannot be obtained even if determination is performed on the entire image. Therefore, in the fourth embodiment of the image type determination unit according to the present invention, as shown in FIG. 8B, the entire image is divided into a plurality of local regions, and the image is determined for each local region.

  FIG. 9 is an explanatory diagram illustrating an image type determination unit 400 according to a fourth embodiment of the image type determination unit.

  Referring to FIG. 9, an image type determination unit 400 according to the fourth example of the image type determination unit has the same configuration as the image type determination unit 100 according to the first example of the image type determination unit. Compared with the image type determination unit 100, the average value calculation unit 430, the relative value calculation unit 440, and the image determination unit 450 are different.

  The frequency band signal detection unit 110 includes first frequency band signal detection means 112 and second frequency band signal detection means 114. Similar to the image type determination unit 100 according to the first embodiment of the image type determination unit, the first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 perform processing for each pixel, Only image signals in the frequency band are allowed to pass.

  The noise removing unit 120 includes first noise removing means 122 and second noise removing means 124. Similar to the image type determining unit 100 according to the first embodiment of the image type determining unit, the first noise removing unit 122 removes noise of the image signal filtered by the first frequency band signal detecting unit 112 for each pixel. The amplitude characteristic value is output for each pixel. The second noise removing unit 124 removes noise from the image signal filtered by the second frequency band signal detecting unit 114 for each pixel.

  The average value calculation unit 430 includes first average value calculation means 432 and second average value calculation means 434. As described above, the first average value calculation unit 132 according to the first embodiment of the image type determination unit receives the amplitude characteristic value for each pixel from which the noise has been removed by the first noise removal unit 122, and the entire image The average value with respect to the absolute value of the amplitude characteristic value at is calculated. On the other hand, the first average value calculation unit 432 according to the fourth embodiment of the image type determination unit receives the amplitude characteristic value for each pixel from which the noise has been removed by the first noise removal unit 122, and FIG. For each local region as shown in (b), an average value for the absolute value of the amplitude characteristic value is calculated.

  Similarly, the second average value calculation unit 434 according to the fourth embodiment of the image type determination unit receives the amplitude characteristic value for each pixel from which noise has been removed by the second noise removal unit 124, and FIG. For each local region as shown in (2), an average value for the absolute value of the amplitude characteristic value is calculated. Needless to say, the local region need not be a region having a uniform size as shown in FIG. 8B, and can be a region having an arbitrary size. Further, the local regions may partially overlap each other.

  The relative value calculation unit 440 calculates a relative value for each local region with respect to the average value for each local region calculated by the average value calculation unit 430. Here, the relative value is calculated in the same manner as the image type determination unit 100 according to the first embodiment of the image type determination unit.

  The image determination unit 450 performs image determination for each local region based on the relative value for each local region calculated by the relative value calculation unit 440, and outputs a determination result for each local region. Therefore, the image processing unit 1300 can perform image processing such as enhancement processing for each local region.

  FIG. 9 shows the same configuration as that of the image type determination unit 100 according to the first example of the image type determination unit. However, the configuration is not limited to the above-described configuration, and the second example of the image type determination unit. The image type determination unit 200 according to the above or the image type determination unit 300 according to the third embodiment of the image type determination unit can be configured. Therefore, it goes without saying that the image type determination unit 400 according to the fourth embodiment of the image type determination unit can determine an image by calculating one or a plurality of relative values according to the configuration. .

  As described above, the image type determination unit 400 according to the fourth embodiment of the image type determination unit uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Then, the image is determined based on the relative value of the average value of the amplitude characteristic values in the middle frequency band with respect to the average value of the amplitude characteristic values in the low frequency band. Further, since the image determination can be performed for each local region, the image determination can be performed accurately even for an image in which an HD resolution image and a pseudo HD resolution image are mixed.

  Therefore, the image type determining unit 400 according to the fourth embodiment of the image type determining unit adds a side panel when an SD resolution image is up-converted as in the conventional image type determining unit. The determination of the image can be performed regardless of the up-conversion method by which the image is up-converted.

  Further, the image type determination unit 400 according to the fourth embodiment of the image type determination unit divides the entire image into local regions and performs image determination for each local region. The image determination can be performed even for an image in which an HD resolution image and a pseudo HD resolution image are mixed.

[Program for image determination]
The program for causing the image type determination unit 400 according to the fourth embodiment of the image type determination unit described above to function as a computer performs image determination regardless of the up-conversion method of the image. be able to. Further, even if an HD resolution image and a pseudo HD resolution image are mixed, image determination can be performed accurately.

[Fifth Example of Image Type Determination Unit]
Next, a description will be given of a fifth embodiment of the image type determination unit according to the present invention, which is another embodiment capable of determining an image for an image in which an HD resolution image and a pseudo HD resolution image are mixed. To do. FIG. 10 is an explanatory diagram illustrating an image type determination unit 500 according to the fifth embodiment of the image type determination unit.

  Referring to FIG. 10, an image type determination unit 500 according to the fifth example of the image type determination unit has the same configuration as the image type determination unit 100 according to the first example of the image type determination unit. The frequency band signal detection unit 110, the noise removal unit 120, and the average value calculation unit 130 have the same functions as the image type determination unit 100. In the first to fourth embodiments of the image type determination unit according to the present invention, the image signal input to the image type determination unit is a signal representing the entire image. An image signal for each local region is input in advance to the image type determination unit 500 according to the fifth example of the type determination unit. With the above configuration, the image type determination unit 500 is different from the image type determination unit 100 in that the image type determination unit 550 is different from the image determination unit 150 according to the first example of the image type determination unit.

  The image determination unit 550 sequentially performs image determination for each local region based on the relative value calculated by the relative value calculation unit 140, that is, the relative value for each local region. Here, the image determination unit 550 starts from the time point when the first frequency band signal detection unit 112 and the second frequency band signal detection unit 114 included in the frequency band signal detection unit 110 detect a signal in a predetermined frequency band. Since the image signal of the local region is processed consistently until the determination is made, the determination of the image of the local region can be performed using only the frequency amplitude characteristics of the image of the local region. Therefore, the image type determination unit 500 according to the fifth example of the image type determination unit does not need to perform processing for dividing the image signal of the entire image into the signal of the local region, and noise or the like is signaled as much as processing is unnecessary. The possibility of getting on can be reduced.

  The image determination unit 550 may include a determination result holding unit that holds a determination result. By providing the determination result holding means, not only the image determination of the local area but also the image determination of the entire image can be performed.

  Note that the configuration of the image type determination unit 500 according to the fifth example of the image type determination unit is not limited to the above. For example, if the configuration shown in FIG. The image type determination unit 500 according to the fifth embodiment can simultaneously determine a plurality of local region images.

  FIG. 10 shows the same configuration as that of the image type determination unit 100 according to the first example of the image type determination unit, but is not limited to the above configuration, and the second example of the image type determination unit. The image type determination unit 200 according to the above or the image type determination unit 300 according to the third embodiment of the image type determination unit can be configured. Therefore, it goes without saying that the image type determination unit 500 according to the fifth embodiment of the image type determination unit can determine an image by calculating one or a plurality of relative values according to the configuration. .

  As described above, the image type determination unit 500 according to the fifth embodiment of the image type determination unit uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Then, the image is determined based on the relative value of the average value of the amplitude characteristic values in the middle frequency band with respect to the average value of the amplitude characteristic values in the low frequency band.

  Further, since the image type determination unit 500 according to the fifth example of the image type determination unit consistently processes the image signal of the local region from the time when the signal in the frequency band is detected until the determination of the image is performed. The determination of the local region image can be performed only with the frequency amplitude characteristics of the local region image. Therefore, it is not necessary to separate the image signal of the entire image into the signal of the local area, and the possibility that noise or the like is applied to the signal can be reduced, so that accurate image determination can be performed.

  Therefore, the image type determination unit 500 according to the fifth example of the image type determination unit adds a side panel when an SD resolution image is up-converted as in the conventional image type determination unit. The determination of the image can be performed regardless of the up-conversion method by which the image is up-converted.

  In addition, since an image signal for each local region is input and image determination can be performed for each local region, even if the image is a mixture of an HD resolution image and a pseudo HD resolution image, Judgment can be made accurately.

[Sixth Example of Image Type Determination Unit]
For example, even if the scenes to be displayed are the same, the television receiving apparatus sequentially rewrites and displays images at a frame rate of 30 frames / second according to the NTSC (National Television Standards Committee) standard. For this reason, for example, it is not preferable that the relative value greatly fluctuates in the same scene, that is, that the determination result of the image greatly differs in the same scene. Therefore, a sixth embodiment of the image type determination unit according to the present invention that can suppress a large variation in the image determination result in a predetermined period will be described next. FIG. 11 is an explanatory diagram illustrating an image type determination unit 600 according to the sixth embodiment of the image type determination unit.

  Referring to FIG. 11, the image type determination unit 600 according to the sixth example of the image type determination unit includes a normalization in addition to the configuration of the image type determination unit 100 according to the first example of the image type determination unit. A processing unit 610 and a time smoothing unit 620 are further provided.

  The frequency band signal detection unit 110, the noise removal unit 120, the average value calculation unit 130, and the relative value calculation unit 140 have the same functions as the image type determination unit 100 according to the first example of the image type determination unit. The relative value calculation unit 140 outputs a relative value used for image determination.

  The normalization processing unit 610 normalizes the relative value calculated by the relative value calculation unit 140, and converts the relative value calculated by the relative value calculation unit 140 into a normalized relative value in a range of 0 ≦ relative value ≦ 1. to correct. The normalization processing unit 610 does not perform normalization processing when the relative value calculated by the relative value calculation unit 140 is a value that does not require normalization. Here, normalization in the normalization processing unit 610 can be performed by setting 1 when the relative value calculated by the relative value calculation unit 140 is 1 or more. It goes without saying that the normalization method is not limited to the above.

  The time smoothing unit 620 smoothes the relative value output from the normalization processing unit 610. Smoothing in the time smoothing unit 620 is performed so as to satisfy the relationship represented by the following Equation 3.

Y _m = (1−α) X _m + αY _m−1 (Expression 3)

In Equation 3, Y _m (m is an integer equal to or greater than 2) represents a smoothed relative value output from the time smoothing unit 620. X _m is a relative value output from the normalization processing unit 610, and Y _m−1 is a relative value output by the time smoothing unit 620 last time. Α (0 <α <1) is a weighting coefficient, and the smoothing ability of the time smoothing unit 620 can be determined by setting α in advance. The time smoothing unit 620 can be configured with a low-pass filter or the like, but is not limited thereto.

  FIG. 11 shows the same configuration as that of the image type determination unit 100 according to the first embodiment of the image type determination unit, but is not limited to the above configuration, and the image type determination unit according to the present invention It can be set as the structure similar to the 2nd-5th Example. Accordingly, the image type determination unit 600 according to the sixth embodiment of the image type determination unit calculates one or a plurality of relative values according to the configuration, smoothes the relative values, and determines an image. It goes without saying that we can do it.

  As described above, the image type determination unit 600 according to the sixth embodiment of the image type determination unit uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Then, the image is determined based on the relative value of the average value of the amplitude characteristic values in the middle frequency band with respect to the average value of the amplitude characteristic values in the low frequency band.

  In addition, the image type determination unit 600 according to the sixth embodiment of the image type determination unit normalizes the relative value, smoothes the relative value based on a predetermined weighting coefficient α, and controls the fluctuation of the relative value. To do.

  Therefore, the image type determination unit 600 according to the sixth embodiment of the image type determination unit adds a side panel when an SD resolution image is up-converted as in the conventional image type determination unit. The determination of the image can be performed regardless of the up-conversion method by which the image is up-converted.

  In addition, the image type determination unit 600 according to the sixth embodiment of the image type determination unit performs time smoothing of the calculated relative value, so that the image determination result greatly changes during a predetermined period. Therefore, accurate image determination can be performed. Therefore, the image processing unit 1300 can uniformly perform image processing such as enhancement processing in a predetermined period.

  In the first to sixth embodiments of the image type determination unit according to the present invention described above, the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image can be used to perform any up-conversion of the image. Regardless of whether the image is up-converted by this method, the image can be accurately determined. Therefore, the image processing apparatus 1000 according to the first embodiment of the present invention including the image type determination unit always performs processing suitable for an image regardless of the up-conversion method of the image. It can be performed.

[Programs related to image processing]
The above-described program for causing the image processing apparatus 1000 according to the first embodiment of the present invention to function as a computer determines an image regardless of the up-conversion method by which the image is up-converted. It is possible to perform processing suitable for the above.

(Second Embodiment)
Next, an image processing apparatus according to the second embodiment of the present invention will be described. FIG. 12 is an explanatory diagram showing an image processing apparatus 2000 according to the second embodiment of the present invention.

  Referring to FIG. 12, an image processing apparatus 2000 according to the second embodiment of the present invention has the same configuration and function as the image processing apparatus 1000 according to the first embodiment of the present invention shown in FIG. Compared with the image processing apparatus 1000, the image processing apparatus 1000 further includes a scene change detection unit 2100, and the image type determination unit 2200 is different from the image type determination unit 1200 according to the first embodiment.

  The scene change detection unit 2100 detects a scene change of an image signal transmitted from a broadcasting station or the like and outputs a scene change signal. Here, the change in scene refers to, for example, a case where an image represented by an image signal transmitted from a broadcasting station or the like changes from a movie to a CM (Commercial Message), or a case where the composition of the image changes. In particular, when an image changes from a movie to a commercial, it is often possible to switch between an HD resolution image and a pseudo HD resolution image. Therefore, it is preferable to detect a change in a scene and process an image suitable for each scene in order to improve image quality.

  The scene change detection unit 2100 can detect a scene change by processing an image signal, an audio signal transmitted together with the image signal from a broadcasting station, or the like. It should be noted that the method for detecting a scene change in the scene change detection unit 2100 is not limited to the above, and it is needless to say that it can be detected using, for example, an attached information signal indicating a scene change. The scene change signal can be a digital signal, but is not limited to the above, and may be an analog signal.

  The image type determination unit 2200 determines an image for each scene based on the detection result of the scene change detection unit 2100. Next, the image type determination unit 2200 will be described.

[Seventh Example of Image Type Determination Unit]
FIG. 13 is an explanatory diagram illustrating an image type determination unit 700 according to the seventh embodiment of the image type determination unit. Referring to FIG. 13, an image type determination unit 700 according to the seventh example of the image type determination unit has the same configuration as the image type determination unit 600 according to the sixth example of the image type determination unit illustrated in FIG. 11. Compared with the image type determination unit 600, the time smoothing unit 710 is different from the time smoothing unit 620 according to the image type determination unit 600.

  The relative value normalized by the normalization processing unit 610 and the scene change signal output from the scene change detection unit 2100 according to the scene change are input to the time smoothing unit 710. When a scene change signal is input, the time smoothing unit 710 newly performs time smoothing of relative values from the input time point. That is, the time smoothing unit 710 does not smooth the relative value continuously before and after the change of the scene, but smoothes the relative value for each scene. Therefore, the image determination unit 150 can determine an image based on a relative value for each scene, and thus can accurately determine an image without being affected by changes in the scene.

  13 shows the same configuration as that of the image type determination unit 100 according to the first embodiment of the image type determination unit. However, the configuration is not limited to the above-described configuration, and the image type determination unit according to the present invention It can be set as the structure similar to the 2nd-5th Example. Therefore, the image type determination unit 700 according to the seventh embodiment of the image type determination unit calculates one or a plurality of relative values according to the configuration, smoothes the relative values, and sets the image for each scene. Needless to say, it can be determined.

  As described above, the image type determination unit 700 according to the seventh embodiment of the image type determination unit uses the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Then, the image is determined based on the relative value of the average value of the amplitude characteristic values in the middle frequency band with respect to the average value of the amplitude characteristic values in the low frequency band.

  In addition, the image type determination unit 700 according to the seventh example of the image type determination unit normalizes the relative value and determines the same as the image type determination unit 600 according to the sixth example of the image type determination unit. The relative value can be smoothed based on the weighting coefficient α, and the fluctuation of the relative value can be controlled.

  Therefore, the image type determination unit 700 according to the seventh embodiment of the image type determination unit adds a side panel when an SD resolution image is up-converted as in the conventional image type determination unit. The determination of the image can be performed regardless of the up-conversion method by which the image is up-converted.

  In addition, the image type determination unit 700 according to the seventh embodiment of the image type determination unit performs time smoothing of the relative value calculated for each scene, so that the image determination result greatly changes in the same scene. Since this can be suppressed, accurate image determination can be performed for each scene.

  Therefore, the image processing apparatus 2000 according to the second embodiment of the present invention including the image type determination unit 700 according to the seventh example of the image type determination unit uses the up-conversion method of the image. Regardless of this, even when an HD resolution image and a pseudo HD resolution image are switched, processing suitable for the image can always be performed according to a change in scene.

[Programs related to image processing]
A program for causing the above-described image processing apparatus 2000 according to the second embodiment of the present invention to function as a computer determines an image regardless of the up-conversion method by which the image is up-converted. In addition, it is possible to perform processing suitable for the image.

  In addition, although the image processing apparatus has been described as the first and second embodiments of the present invention, the first and second embodiments of the present invention are not limited to such a form, and may be a television receiver, a PDA ( The present invention can be applied to computers such as Personal Digital Assistants (UMPC) and Ultra Mobile Personal Computer (UMPC), and portable communication devices such as mobile phones and PHS (Personal Handyphone System).

  By applying the first and second embodiments of the present invention to a television receiver, a computer, a portable communication device, etc., for example, when receiving a high-definition broadcast or a one-segment broadcast, the received image signal Since suitable processing can be performed, the image quality can be improved.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to the example which concerns. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Understood.

  For example, in the first to seventh embodiments according to the image type determination unit of the present invention, an average value calculation unit that calculates an average value is provided. A median calculation unit can also be used. Even if the relative value is calculated using the median value instead of the average value, it is possible to use the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG. Therefore, the image type determination unit can determine the image.

  In the first to fifth embodiments relating to the image type determination unit of the present invention, a normalization processing unit 610 included in the sixth and seventh embodiments of the image type determination unit of the present invention is provided. However, the present invention is not limited to such a configuration, and the first to fifth embodiments according to the image type determination unit of the present invention may include a normalization processing unit. As described above, the normalization processing unit always corrects the relative value to the normalized relative value in the range of 0 ≦ relative value ≦ 1, so the image type determination unit including the normalization processing unit is shown in FIG. The image can be determined using the difference in frequency amplitude characteristics between the HD resolution image and the pseudo HD resolution image as shown in FIG.

  Further, in the first to seventh embodiments relating to the image type determination unit of the present invention, the configuration including the noise removing unit has been described. However, the configuration is not limited thereto, and a configuration not including the noise removing unit may be employed. it can. As described above, the image type determination unit according to the present invention determines an image using a difference in frequency amplitude characteristics between an HD resolution image and a pseudo HD resolution image as shown in FIG. Even if the image is affected by noise, the image can be determined regardless of the up-conversion method of the image.

  The configuration described above can be easily changed by those skilled in the art, and should be understood as belonging to the equivalent scope of the present invention.

It is explanatory drawing which shows the image of HD resolution, and the image which upconverted the image of SD resolution. FIG. 2 is an explanatory diagram illustrating a difference in frequency amplitude characteristics between an HD resolution image illustrated in FIG. 1A and a pseudo HD resolution image illustrated in FIG. It is explanatory drawing which shows the image processing apparatus which concerns on the 1st Embodiment of this invention. 3 is a flowchart illustrating an image processing method according to an embodiment of the present invention. It is explanatory drawing which shows the image type determination part which concerns on 1st Example of an image type determination part. It is explanatory drawing which shows the image type determination part which concerns on 2nd Example of an image type determination part. It is explanatory drawing which shows the image type determination part which concerns on 3rd Example of an image type determination part. It is explanatory drawing which shows the image in which the image of HD resolution and the image of pseudo HD resolution are mixed. It is explanatory drawing which shows the image type determination part which concerns on the 4th Example of an image type determination part. It is explanatory drawing which shows the image type determination part which concerns on the 5th Example of an image type determination part. It is explanatory drawing which shows the image type determination part which concerns on the 6th Example of an image type determination part. It is explanatory drawing which shows the image processing apparatus which concerns on the 2nd Embodiment of this invention. It is explanatory drawing which shows the image type determination part which concerns on the 7th Example of an image type determination part.

Explanation of symbols

100, 200, 300, 400, 500, 600, 700 Image type determination unit 110, 210 Frequency band signal detection unit 120, 220 Noise removal unit 130, 230, 430 Average value calculation unit 140, 240, 340, 440 Relative value calculation Unit 150, 250, 450, 550 Image determination unit 620, 710 Time smoothing unit 1000, 2000 Image processing device 1200, 2200 Image type determination unit 1300 Image processing unit 2100 Scene change detection unit

Claims (10)

A frequency band signal detector for detecting signals of a plurality of frequency bands from the image signal;
An average value calculating unit that calculates an average value for a characteristic value corresponding to an amplitude for each signal in a plurality of frequency bands detected by the frequency band signal detecting unit;
Among the average values calculated for each of the plurality of frequency bands, a relative value calculation unit that calculates a relative value of the other average value with respect to one of the average values;
An image determination unit that determines an image based on the relative value calculated by the relative value calculation unit;
An image type determination unit having
An image processing unit that processes an image based on an image determination result of the image type determination unit;
An image processing apparatus comprising: The image type determination unit further includes a noise removal unit that removes the signal having the characteristic value less than a predetermined threshold for each of the signals of the plurality of frequency bands detected by the frequency band signal detection unit,
The image processing apparatus according to claim 1, wherein the average value calculation unit calculates an average value for the characteristic value from which noise has been removed by the noise removal unit.   The relative value calculation unit is based on the average value calculated in a predetermined frequency band determined in advance and the average value calculated in a frequency band higher than the predetermined frequency band determined in advance, The image processing apparatus according to claim 1, wherein the relative value is calculated.   The lower limit frequency of the predetermined frequency band is lower than the theoretically effective upper limit frequency according to the up-conversion method with respect to the Nyquist frequency of the image up-converted from the standard resolution image, and the high frequency band The lower limit frequency of the frequency band is higher than the theoretically effective upper limit frequency according to the up-conversion method with respect to the Nyquist frequency of the image up-converted from the standard resolution image. Image processing apparatus.   The relative value calculation unit includes the average value calculated in a predetermined frequency band determined in advance and the average value calculated in a plurality of frequency bands higher than the predetermined frequency band determined in advance. The image processing apparatus according to claim 1, wherein a plurality of the relative values are calculated based on the information.   The relative value calculation unit is configured to calculate the average value calculated in a plurality of predetermined predetermined frequency bands and the average calculated in a frequency band higher than each of the predetermined predetermined frequency bands. The image processing apparatus according to claim 1, wherein a plurality of the relative values are calculated based on the value.   The average value calculation unit calculates the average value in each frequency band of the plurality of frequency bands for each local region of the entire image, and the image determination unit calculates the local value for each local region based on the relative value. The image processing apparatus according to claim 1, wherein an image is determined.   The image type determination unit further includes a time smoothing unit that temporally smoothes the relative value, and the image determination unit determines an image based on the time-smoothed relative value. The image processing apparatus according to 1. A scene change detection unit for detecting a change in the scene of the image represented by the image signal;
The image processing according to claim 8, wherein the time smoothing unit performs time smoothing of the relative value from a point in time when a scene change is detected by the scene change detection unit. apparatus. Detecting signals in a plurality of frequency bands from the image signal;
Calculating an average value for a characteristic value corresponding to an amplitude for each of the detected signals of the plurality of frequency bands;
Calculating a relative value of another average value with respect to one of the average values calculated for each of the plurality of frequency bands;
Determining an image based on the calculated relative value;
Processing the image based on the determination result of the image;
An image processing method characterized by comprising:

Images