JP4111108B2 - Image processing apparatus and method - Google Patents

Description

  The present invention relates to an image processing apparatus and method suitable for use when, for example, converting a standard resolution or low resolution image into an arbitrary high resolution image.

  For example, a standard resolution or low resolution image (hereinafter referred to as “SD (Standard Definition) image” as appropriate) is converted into a high resolution image (hereinafter referred to as “HD (High Definition) image” as appropriate), or an image is displayed. In the case of enlarging the pixel value, interpolation (compensation) of the missing pixel value is performed by a so-called interpolation filter or the like.

  However, even if pixel interpolation is performed using an interpolation filter, it is difficult to restore a high-resolution image because the HD image component (high frequency component) that is not included in the SD image cannot be restored. .

  Therefore, the applicant of the present application has previously proposed an image conversion device that converts an SD image into an HD image that also includes a high frequency component that is not included therein (see, for example, Patent Documents 1 to 3 below).

JP-A-7-193789 JP-A-11-55630 JP 2000-134586 A

  By the way, in this image conversion apparatus, the SD image can be converted into an HD image of an arbitrary high resolution by changing the resolution axis intensity. However, if the resolution axis strength is increased too much, the frequency characteristics are emphasized from the middle range, and preshoot / overshoot occurs in the output signal. Further, since the high frequency component included in the input signal is emphasized, the degree to which the resolution axis strength is increased is limited from the viewpoint of S / N (Signal / Noise).

  The present invention has been proposed in view of such a conventional situation, and an image processing apparatus and method that does not cause problems in terms of S / N and preshoot / overshoot even when the resolution axis strength is increased. The purpose is to provide.

In order to achieve the above-described object, an image processing apparatus according to the present invention includes a frequency characteristic adjusting unit that adjusts frequency characteristics of a plurality of types of input first image signals, and a frequency characteristic that is adjusted by the frequency characteristic adjusting unit. Image conversion means for converting the adjusted first image signal into a second image signal having a larger number of pixels than the first image signal, and the frequency characteristic adjustment means is provided by the image conversion means. A frequency characteristic of the first image signal is attenuated as a high frequency range so as to show a characteristic of a substantially low-pass filter that is substantially opposite to a frequency characteristic change. As the image signal is narrower and the information amount of the high-frequency component is smaller, the attenuation amount of the frequency characteristic is increased .

In order to achieve the above-described object, an image processing method according to the present invention includes a frequency characteristic adjustment step for adjusting frequency characteristics of a plurality of types of input first image signals, and the frequency characteristic adjustment step. An image conversion step of converting the first image signal having a changed frequency characteristic into a second image signal having a larger number of pixels than the first image signal. In the frequency characteristic adjustment step, The frequency characteristics of the first image signal are attenuated as the frequency increases so that the characteristics of the low-pass filter, which is substantially the reverse characteristic of the frequency characteristic change in the image conversion process, are shown, and the plurality of types of first image signals Among them, the attenuation amount of the frequency characteristic is increased as the image signal has a narrower band and a smaller amount of high-frequency component information .

  In such an image processing apparatus and method, when the input first image signal is converted into a second image signal having a larger number of pixels than the first image signal, the first image signal is converted into the first image signal before the first image signal. The frequency characteristic of the first image signal is attenuated as the frequency increases so that the frequency characteristic of the image signal of FIG.

  According to the image processing device and the method thereof according to the present invention, when the input first image signal is converted into the second image signal having a larger number of pixels than the first image signal, As the frequency characteristics of the first image signal substantially indicate the characteristics of a low-pass filter, the frequency characteristics of the first image signal are attenuated as the frequency increases, thereby causing problems of S / N and preshoot / overshoot. It is possible to obtain a second image having an arbitrary high resolution from the first image signal.

  Hereinafter, specific embodiments to which the present invention is applied will be described in detail with reference to the drawings. In this embodiment, a standard resolution or low resolution image (hereinafter referred to as an “SD (Standard Definition) image”) composed of an input video signal is used as an arbitrary high resolution image. The image processing apparatus is appropriately converted to an “HD (High Definition) image” and further converted into the number of pixels of the display and displayed.

  First, FIG. 1 shows a schematic configuration of an image processing apparatus according to the first embodiment. As shown in FIG. 1, an image processing apparatus 1 according to the present embodiment includes a tuner 10, an image processing unit 11, an image conversion unit 20, a pixel number conversion unit 21, a Y system processing unit 22, and a C system. The processing unit 23, a matrix circuit 24, a drive / cutoff adjustment unit 25, and a display display 26 are included.

  The image processing unit 11 includes a Y / C separation circuit 12, a chroma demodulation circuit 13, a selector 14, a frequency adjustment circuit 15, and an amplitude level adjustment circuit 16. The tuner 10 uses a CVBS (Composite Video Burst Signal). Various video signals such as an RF signal demodulated into a signal, a CVBS signal from an external video device, a YUV signal, and an S (Y and C) signal are input.

  The Y / C separation circuit 12 decodes the CVBS signal and the S signal into a YC signal, supplies the Y signal to the selector 14, and supplies the C signal to the chroma demodulation circuit 13. The chroma demodulation circuit 13 demodulates the C signal into a U signal and a V signal, and supplies them to the selector 14 respectively. The selector 14 supplies the Y signal to the frequency adjustment circuit 15 among the CVBS signal and S signal decoded into the YUV signal and the YUV signal whose input is the YUV signal, and adjusts the amplitude level of the U signal and the V signal. Supply to circuit 16. The frequency adjustment circuit 15 adjusts the frequency characteristics of the Y signal as described later in accordance with the resolution axis strength in the subsequent image conversion unit 20 and supplies the adjusted Y signal to the amplitude level adjustment circuit 16. Then, the amplitude level adjustment circuit 16 adjusts the amplitude level of the YUV signal and outputs it as a 525i signal (interlaced signal having 525 lines).

  The image conversion unit 20 converts the 525i signal into, for example, a 1050i signal having twice the number of lines by performing an adaptive process for obtaining a prediction value of the HD image by linear combination of the SD image and a predetermined prediction coefficient. At this time, the image conversion unit 20 can convert the image into an arbitrary high-resolution HD image by changing the resolution axis strength. The image conversion unit 20 supplies the converted YUV signal to the pixel number conversion unit 21. The image conversion process in the image conversion unit 20 will be described in detail later.

  The pixel number conversion unit 21 converts the video format (number of pixels in one screen) of the YUV signal supplied from the image conversion unit 20 so as to match the number of pixels of the display 26. The pixel number conversion unit 21 supplies the Y signal to the Y-system processing unit 22 and supplies the UV signal to the C-system processing unit 23 among the YUV signals having undergone pixel number conversion.

  The Y-system processing unit 22 includes a DC transmission rate circuit, a black expansion circuit, a sharpness circuit (all not shown), and performs predetermined processing on the Y signal. On the other hand, the C-system processing unit 23 includes a color level adjustment circuit (not shown) and the like, and performs predetermined processing on the UV signal. The Y-system processing unit 22 and the C-system processing unit 23 supply the processed Y signal and UV signal to the matrix circuit 24, respectively.

  The matrix circuit 24 converts the YUV signal supplied from the Y system processing unit 22 and the C system processing unit 23 into an RGB signal, and supplies the converted RGB signal to the drive / cutoff adjustment unit 25. Then, the drive / cutoff adjustment unit 25 adjusts the drive level and the cut-off level for the display display 26 and displays them on the display display 26.

  Here, in the image conversion unit 20 described above, high-frequency components not included in the SD image are performed by performing an adaptive process for obtaining a prediction value of the HD image by linear combination of the SD image and a predetermined prediction coefficient. Has been made to be restored.

That is, for example, the predicted value E [y] of the pixel value y of a pixel constituting the HD image (hereinafter referred to as “HD pixel” as appropriate) is used as the pixel value of several SD pixels (pixels constituting the SD image). (hereinafter, appropriately referred to as "learning _data".) x _{1, x} 2, determined ... and predetermined prediction coefficients _{w 1, w} 2, the linear combination model defined by a linear combination of the ... Think about it. In this case, the predicted value E [y] can be expressed by the following formula (1).

  Therefore, in order to generalize, a matrix W composed of a set of prediction coefficients w, a matrix X composed of a set of learning data, and a matrix Y ′ composed of a set of predicted values E [y] are expressed by the following equations (2) to ( When defined in 4), the observation equation as shown in equation (5) is established.

  Then, it is considered to apply the least square method to this observation equation to obtain a predicted value E [y] close to the pixel value y of the HD pixel. In this case, a matrix Y composed of a set of true pixel values y of HD pixels serving as teacher data and a matrix E composed of a set of residuals e of predicted values E [y] with respect to the pixel values y of HD pixels are expressed by the following equations. When defined by (6) and (7), a residual equation such as expression (8) is established from the above expression (5).

In this case, the prediction coefficient w _i for obtaining the predicted value E [y] close to the pixel value y of the HD pixel can be obtained by minimizing the square error as shown in the equation (9), for example.

Therefore, when the above-mentioned square error differentiated by the prediction coefficient w _i is 0, that is, the prediction coefficient w _i satisfying the following expression (10) is the prediction value E [y near the pixel value y of the HD pixel. ] Is the optimum value.

Therefore, first, the following equation (11) is established by differentiating the equation (8) by the prediction coefficient w _i .

  Then, from the equations (10) and (11), the following equation (12) is obtained.

  Further, considering the relationship among the learning data x, the prediction coefficient w, the teacher data y, and the residual e in the residual equation of the equation (8), the normality as shown in the following equation (13) is obtained from the equation (12). An equation can be obtained.

  The normal equation of the equation (13) can be established by the same number as the number of prediction coefficients w to be obtained. Therefore, by solving the equation (13) (however, to solve the equation (13), the equation In (13), the matrix composed of the coefficients related to the prediction coefficient w needs to be regular), and the optimal prediction coefficient w can be obtained. In solving equation (13), for example, a sweep-out method (Gauss-Jordan elimination method) or the like can be applied.

  As described above, the optimum prediction coefficient w is obtained, and further, using the prediction coefficient w, it is adaptive to obtain the prediction value E [y] close to the pixel value y of the HD pixel by the equation (1). It is processing.

  Note that the adaptive processing is different from the interpolation processing in that the components included in the HD image that are not included in the SD image are reproduced. In other words, the adaptive process is the same as the interpolation process using a so-called interpolation filter as far as Expression (1) is seen, but the prediction coefficient w corresponding to the tap coefficient of the interpolation filter is determined using the teacher data y. In other words, since it is obtained by learning, the components included in the HD image can be reproduced. That is, a high-resolution image can be easily obtained. From this, it can be said that the adaptive process is a process having an effect of creating an image (its resolution).

  FIG. 2 shows an internal configuration example of an image conversion unit that converts an SD image into an HD image by the adaptive processing as described above.

  The SD image is supplied to the class classification unit 101 and the adaptive processing unit 104. The class classification unit 101 includes a class tap generation circuit 102 and a class classification circuit 103, where an HD pixel (a focused HD pixel) for which a prediction value is to be obtained by adaptive processing (hereinafter referred to as a “target pixel” as appropriate). Is classified into a predetermined class based on the property of the pixel of the SD image corresponding to the target pixel.

  That is, the class tap generation circuit 102 is used for classifying the target pixel. As an SD pixel corresponding to the target pixel (hereinafter referred to as “class tap” as appropriate), for example, a plurality of SD pixels having a predetermined positional relationship with the target pixel are supplied to the class classification unit 101. And is supplied to the class classification circuit 103. The class classification circuit 103 detects the pixel value pattern (pixel value distribution) of the SD pixels constituting the class tap from the class tap generation circuit 102, and the value assigned in advance to the pattern is used as the class of the target pixel. To the adaptive processing unit 104.

Specifically, for example, when an HD image is composed of pixels (HD pixels) indicated by x in FIG. 3, and an SD image is composed of pixels (SD pixels) indicated by circles in FIG. To do. In FIG. 3, the SD image is configured so that the number of horizontal or vertical pixels of the HD image is halved. Here, in FIG. 3 (the same applies to FIGS. 4 to 6 described later), the (i + 1) th SD pixel from the left and the (j + 1) th SD pixel from the top (indicated by a circle in the figure) is represented as X _{i, j.} Similarly, the i ′ + 1-th HD pixel from the left and the j ′ + 1-th HD pixel from the top (portion indicated by a cross in the figure) is represented as Y _{i ′, j ′} . In this case, the position of the SD pixel X _{i, j} matches the position of the HD pixel Y _{2i, 2j} .

Now, for example, if an HD pixel Y _4,4 that coincides with the position of X _2,2 is a pixel of interest as an SD pixel, the class tap generation circuit 102 uses the SD pixel corresponding to the HD pixel Y _4,4. as, for example, SD HD pixels _{Y 4, 4} and 3 × 3 centered on the SD pixels _{X 2, 2} correlation coincides with the position of HD pixel _{Y 4, 4} which are expected to be high for (horizontal × vertical) Pixels X _1,1 , X _2,1 , X _3,1 , X _1,2 , X _2,2 , X _3,2 , X _1,3 , X _2,3 , X _3,3 (dotted line in FIG. 3) SD pixels in a range surrounded by) are extracted, and are used as class taps of the _target pixel (HD pixel) _Y4,4 .

Also, for example, when the HD pixel Y _5,4 right next to the HD pixel Y _4,4 that coincides with the position of X _2,2 is the target pixel, the class tap generation circuit 102 causes the HD pixel Y _As SD pixels corresponding to ₅ and ₄ , for example, as shown by being surrounded by a dotted line in FIG. 4, SD pixels X _{1 and 1} in the class tap formed when the HD pixel Y _{4 and 4} is the target pixel are shown. Instead of ₂ pixels, those including SD pixels X ₄ and ₂ are extracted, and the 9 SD pixels are used as class taps of the _target pixel (HD pixel) Y _{5 and 4} .

Further, for example, when the HD pixel Y _4,5 adjacent below the HD pixel Y _4,4 coinciding with the position of X _2,2 is set as the target pixel, the class tap generation circuit 102 causes the HD pixel As an SD pixel corresponding to Y _4,5 , for example, as shown by being surrounded by a dotted line in FIG. 5, the SD pixel X ₂ in the class tap formed when the HD pixel Y _4,4 is set as the target pixel. ₁ and SD pixels X ₂ and ₄ are extracted, and the nine SD pixels are used as class taps for the _target pixel (HD pixel) Y _4,5 .

For example, when the HD pixel Y _5,5 adjacent to the lower right of the HD pixel Y _4,4 that coincides with the position of X _2,2 is the target pixel, the class tap generation circuit 102 As the SD pixels corresponding to the HD pixels Y ₅ and ₅ , for example, as shown by being surrounded by a dotted line in FIG. 6, the SD pixels in the class tap formed when the HD pixels Y _{4 and 4} are the target pixels. Instead of X _1,1 , those including SD pixels X _4,4 are extracted, and the nine SD pixels are used as class taps of _target pixels (HD pixels) Y _5,5 .

  The class classification circuit 203 detects a pattern of nine SD pixels (pixel values) as the class tap configured by the class tap generation circuit 102, and outputs a value corresponding to the pattern as a class of the target pixel. Is done.

  This class is supplied to an address terminal (AD) of a coefficient ROM (Read Only Memory) 107 in the adaptive processing unit 104.

Here, 8 bits or the like are generally assigned to the pixels constituting the image. Now, assuming that 8 bits are allocated to the SD pixel, for example, even if only the 3 × 3 pixel square class tap shown in FIG. 3 is considered, the number of pixel value patterns is (2 ⁸ ) ^9. The number of streets becomes enormous, and it is difficult to speed up subsequent processing.

  Therefore, as preprocessing before class classification, the class tap is subjected to, for example, ADRC (Adaptive Dynamic Range Coding) processing, which is processing for reducing the number of bits of SD pixels constituting the class tap.

That is, in the ADRC process, first, among the nine SD pixels constituting the class tap, the maximum value (hereinafter referred to as “maximum pixel” as appropriate) and the minimum value (hereinafter referred to as “minimum pixel” as appropriate). Is detected). Then, a difference DR (= MAX−MIN) between the pixel value MAX of the maximum pixel and the pixel value MIN of the minimum pixel is calculated, and this DR is set as a local dynamic range of the class tap. Based on the dynamic range DR, Each pixel value constituting the class tap is requantized to K bits smaller than the original number of assigned bits. That is, the pixel value MIN of the minimum pixel from each pixel value is subtracted forming the class taps, each subtraction value is divided by DR / 2 ^K.

As a result, each pixel value constituting the class tap is expressed by K bits. Accordingly, for example, when K = 1, the number of patterns of the pixel values of the nine SD pixels is (2 ¹ ) ^nine , and the number of patterns is very small compared to the case where ADRC processing is not performed. It can be.

  On the other hand, the adaptive processing unit 104 includes a prediction tap generation circuit 105, a prediction calculation circuit 106, and a coefficient ROM 107, where adaptive processing is performed.

  In other words, in the prediction tap generation circuit 105, a plurality of pixels having a predetermined positional relationship with respect to the target pixel, which are used for obtaining the predicted value of the target pixel in the prediction calculation circuit 106 from the SD image supplied to the adaptive processing unit 104. SD pixels are extracted and supplied to the prediction calculation circuit 106 as prediction taps.

Specifically, for example, HD pixel _{Y 4, 4} is a pixel of interest, when the class tap as described in FIG. 3 constituted, the prediction tap generating circuit 105, for example, an HD pixel _{Y 4, 4} 5 × 5 SD pixels centered on the SD pixel X _2,2 that coincides with the position of the target pixel Y _{4,4 in} the range surrounded by the solid line in FIG. Are extracted and used as prediction taps for the target pixel (HD pixel) _Y4,4 .

Further, for example, when the HD pixel _{Y 5,4} is a pixel of interest, the prediction tap generating circuit 105, for example, as shown enclosed by a solid line in FIG. 4, HD pixel _{Y 4, 4} and the pixel of interest In the prediction tap formed in this case, SD pixels including X pixels ₅ and ₂ are extracted in place of the SD pixels X ₀ and ₂ , and the 25 SD pixels are extracted as pixels of interest (HD pixels). Y _{5 and 4} are class taps.

Further, for example, when the HD pixel Y _4,5 is set as the target pixel, the prediction tap generation circuit 105 sets the HD pixel Y _{4,4 as} the target pixel as shown by a solid line in FIG. In the prediction tap formed in this case, SD pixels X _2,5 are extracted instead of SD pixels X _2,0 , and the 25 SD pixels are extracted as the target pixel (HD pixel). The prediction tap is Y _4,5 .

For example, when the HD pixels Y _{5 and 5} are the target pixels, the prediction tap generation circuit 105 determines that the HD pixels Y _{4 and 4} are the target pixels as indicated by a solid line in FIG. In the class tap formed in this case, the SD pixel X _0,0 including the SD pixel X _5,5 is extracted instead of the SD pixel X _0,0 , and the 25 SD pixels are extracted as the _target pixel (HD pixel). The prediction tap is Y _5,5 .

  In addition to the prediction tap supplied from the prediction tap generation circuit 105, the prediction calculation circuit 106 is also supplied with the prediction coefficient from the coefficient ROM 107.

  That is, the coefficient ROM 107 stores the prediction coefficient obtained by learning in advance for each class, and when a class is supplied from the class classification circuit 103, the coefficient ROM 107 stores it in an address corresponding to the class. The prediction coefficient is read and supplied to the prediction calculation circuit 106.

As a result, the prediction calculation circuit 106 is supplied with the prediction tap corresponding to the target pixel and the prediction coefficient for the class of the target pixel. In the prediction arithmetic circuit 106, prediction coefficients w, w ₂ ,... From the coefficient ROM 107 and prediction taps (constituting SD pixels) x ₁ , x ₂ ,. Is used to calculate the predicted value E [y] of the pixel of interest (HD pixel) y, and this is output as the pixel value of the HD pixel.

  The above processing is performed with all the HD pixels as the target pixel, whereby the SD image is converted into an HD image. Note that the class tap generation circuit 102 and the prediction tap generation circuit 105 perform processing using the same HD pixel as the pixel of interest.

  Next, FIG. 7 shows a configuration example of a learning apparatus that performs a learning process for calculating a prediction coefficient to be stored in the coefficient ROM 107 of FIG.

  The HD image to be the teacher data y in learning is supplied to the thinning circuit 201 and the teacher data extraction circuit 204. In the thinning circuit 201, for example, the number of pixels of the HD image is thinned out. Thus, an SD image is obtained. That is, in the thinning circuit 201, the number of horizontal or vertical pixels of the HD image is halved, whereby an SD image is formed. This SD image is supplied to the class classification unit 202 and the prediction tap generation circuit 203.

  In the class classification unit 202 or the prediction tap generation circuit 203, the same processing as in the class classification unit 101 or the prediction tap generation circuit 105 in FIG. 2 is performed, and thereby the class or prediction tap of the pixel of interest is output, respectively. . The class output from the class classification unit 202 is supplied to the address terminals (AD) of the prediction tap memory 205 and the teacher data memory 206, and the prediction tap output from the prediction tap generation circuit 203 is supplied to the prediction tap memory 205. Note that the class classification unit 202 and the prediction tap generation circuit 203 perform processing using the same HD pixel as the target pixel.

  In the prediction tap memory 205, the prediction tap supplied from the prediction tap generation circuit 203 is stored at the address corresponding to the class supplied from the class classification unit 202.

  On the other hand, in the teacher data extraction circuit 204, the HD pixel that is the target pixel in the class classification unit 202 and the prediction tap generation circuit 203 is extracted from the HD image supplied thereto and supplied to the teacher data memory 206 as teacher data. Is done.

  The teacher data memory 206 stores the teacher data supplied from the teacher data extraction circuit 204 at an address corresponding to the class supplied from the class classification unit 202.

  The above processing is sequentially performed with all HD pixels constituting all HD images prepared for learning in advance as the target pixel.

  As a result, the same address in the teacher data memory 206 or the prediction tap memory 205 is located at the position where the HD pixel of the class corresponding to the address or the prediction tap described with reference to FIGS. SD pixels are stored as teacher data y or learning data x, respectively.

  Note that the prediction tap memory 205 and the teacher data memory 206 can store a plurality of pieces of information at the same address, so that a plurality of learnings classified into the same class can be stored at the same address. Data x and teacher data y can be stored.

  Thereafter, the arithmetic circuit 207 reads out the prediction tap or the HD pixel as the teacher data as the learning data stored at the same address from the prediction tap memory 205 or the teacher data memory 206, and uses them, for example, the minimum self A prediction coefficient that minimizes an error between the predicted value and the teacher data is calculated by multiplication. That is, in the arithmetic circuit 207, the normal equation shown in Expression (13) is established for each class, and the prediction coefficient is obtained by solving this.

  As described above, the prediction coefficient for each class obtained by the arithmetic circuit 207 is stored at an address corresponding to each class in the coefficient ROM 107 of FIG.

  In the learning process as described above, there may occur a class in which the number of normal equations necessary for obtaining the prediction coefficient cannot be obtained. For such a class, for example, the class is ignored and the normal equation is ignored. The prediction coefficient obtained by solving the above is used as a default prediction coefficient.

  Here, the process in the image conversion unit 20 described above performs an adaptive process for obtaining a predicted value of the HD image by linear combination of the SD image and a predetermined prediction coefficient, so that a high frequency that is not included in the SD image is obtained. Since the component is restored, the output signal from the image conversion unit 20 has a higher frequency component emphasized than the input signal. That is, the image conversion unit 20 has a frequency characteristic that increases as the high frequency component increases. As a result, as schematically shown in FIG. 8, even if the input signal of the image conversion unit 20 is flat, the output signal from the image conversion unit 20 is emphasized as the high frequency component is increased. .

  However, when the high frequency component is emphasized in this way, the preshoot / overshoot of the output signal is conspicuous and the S / N is deteriorated, which may cause deterioration in image quality.

  Therefore, in the image processing apparatus 1 according to the present embodiment, the frequency characteristics in the image conversion unit 20 are taken into consideration, and the image processing unit 11 (specifically, the frequency adjustment circuit 15) in the preceding stage shows a specific reverse characteristic. Specifically, the frequency characteristics are adjusted so as to show the characteristics of an LPF (Low-Pass filter). As a result, as schematically shown in FIG. 9A, when the input signal of the image processing unit 11 is flat, the output signal from the image processing unit 11 (the input signal of the image conversion unit 20) is The higher frequency components are attenuated, and the output signal from the image conversion unit 20 is flat.

  At this time, compared with the S / N of the input signal of the image processing unit 11, the S / N of the output signal is improved by the amount of attenuation of the frequency characteristic of the high frequency component. Further, since the frequency component newly created by the image conversion unit 20 contains almost no random pattern signal, the S / N improvement is ensured even in the output signal from the image conversion unit 20. Is done. That is, as shown in FIG. 10, in the frequency adjustment circuit 15 of the image processing unit 11, the S / N of the portion indicated by the hatching in the figure is improved by setting the frequency characteristics so that the higher frequency component attenuates. It becomes.

  In order to flatten the frequency characteristic of the output of the image conversion unit 20 in FIG. 9A, a frequency characteristic that compensates for the frequency characteristic of the image conversion unit 20 is required for the image processing unit 11 in the previous stage. However, strictly speaking, it is difficult for the image processing unit 11 in the previous stage to compensate the frequency characteristics up to a high frequency range created by the image conversion unit 20. Therefore, as a practical method, a filter having a frequency characteristic as shown in FIG. 9B may be used in the image processing unit 11 and substituted.

  Furthermore, by significantly increasing the attenuation amount of the high frequency component by the frequency adjustment circuit 15 (for example, near 6 dB), it is possible to reduce the ringing of the transient part included in the input signal of the image conversion unit 20. That is, generally, in the band compression / decompression process, the bit distribution in the transient part is less than that in the flat part. When the image conversion unit 20 converts an SD image into an HD image, it has the same property. If the frequency characteristics of the high frequency component are significantly attenuated, the information of the high frequency component falls uniformly. The degree of drop in the output signal from the conversion unit 20 is larger in the transient part due to the bit distribution. Therefore, by attenuating the frequency characteristic in the frequency adjustment circuit 15 as the high frequency component in this way, it is possible to reduce the ringing in the transient portion that is likely to occur in the RF signal to a considerable extent.

  In addition, not only ringing, but also mosquito noise (a type of noise generated during MPEG (Moving Picture Experts Group) decoding) generated around signal waveforms with medium amplitude or higher (especially character signals) can be similarly reduced. It is. Note that this reduction effect increases as the frequency characteristic of the high frequency component is attenuated in the frequency adjustment circuit 15.

  Next, the image processing apparatus 2 shown in FIG. 11 as the second embodiment has the same basic structure as that of the image processing apparatus 1 shown in FIG. 1 except that the S / N detection circuit 17, the microcomputer 27, And the amount of attenuation of the frequency characteristic in the frequency adjustment circuit 15 is variable in accordance with the band of the video signal and the S / N. Therefore, the same components as those of the image processing apparatus 1 previously shown in FIG.

  In the image processing apparatus 2 shown in FIG. 11, the frequency adjustment circuit 15 attenuates the frequency characteristic as the high frequency component is the same as in the first embodiment. A video signal with a smaller amount of information is considered to have less adverse effects due to attenuation of the frequency characteristics. Therefore, the microcomputer 27 according to the present embodiment uses the frequency characteristics in the frequency adjustment circuit 15 in the order of 1) S signal and YUV signal, 2) CVBS signal, and 3) RF signal based on the type of the input video signal. The attenuation is controlled to be large. Specifically, for example, the wide band S signal and YUV signal are attenuated by about 2 dB, the CVBS signal is attenuated by about 2 to 3 dB, and the narrow band RF signal is attenuated by about 6 dB. At this time, the image conversion unit 20 increases the resolution axis intensity by the attenuation amount of the frequency characteristic.

  Further, as described above, the S / N of the video signal can be improved by increasing the attenuation amount of the high-frequency component by the frequency adjustment circuit 15 and increasing the resolution axis strength by the image conversion unit 20. If the S / N of the video signal is poor, it is considered effective to increase the attenuation amount of the frequency characteristic in the frequency adjustment circuit 15 and increase the resolution axis strength in the image conversion unit 20. Therefore, the S / N detection circuit 17 in the present embodiment detects the S / N of the Y signal supplied from the selector 14 and supplies the detection result to the microcomputer 27. The microcomputer 27 controls the adjustment of the attenuation amount of the frequency characteristic in the frequency adjustment circuit 15 according to the detection result of the S / N. For example, the RF signal often has a poor S / N, but by increasing the attenuation amount of the frequency characteristic in the frequency adjustment circuit 15 and increasing the resolution axis strength in the image conversion unit 20 as described above, the S / N can be reduced. Significant improvement is possible.

  Subsequently, the image processing apparatus 3 shown in FIG. 12 as the third embodiment has the same basic structure as the image processing apparatus 1 shown in FIG. 1, but the frequency adjustment circuit 15 includes a coring circuit 18. It is characterized in that the frequency characteristics of a minute amplitude signal having a threshold value Th or less are allowed to pass through without being attenuated. Therefore, the same components as those of the image processing apparatus 1 previously shown in FIG.

  Here, in the first embodiment described above, since the frequency characteristics of the minute amplitude signal are uniformly attenuated, there is a possibility that the detail of the signal may be lost.

  Therefore, in the present embodiment, the coring circuit 18 is built in the frequency adjustment circuit 15 so that the frequency characteristics of the minute amplitude signal are not attenuated. Specifically, the coring circuit 18 exhibits input / output characteristics as shown in FIG. As can be seen from FIG. 13, in the coring circuit 18, the output of a minute amplitude signal that is equal to or less than the threshold Th is zero.

  An example of the internal configuration of the frequency adjustment circuit 15 is shown in FIG. 14A, and an example of a signal waveform at each point of the frequency adjustment circuit 15 is shown in FIG. As shown in FIG. 14A, the frequency adjustment circuit 15 includes delay circuits 30 and 31, multipliers 32, 33, and 34, adders 35 and 36, and the coring circuit 18 described above. The input signal of the frequency adjustment circuit 15 indicated by the signal waveform at the point a is delayed by the delay amount td by the delay circuit 30 to become the signal waveform at the point b, and further delayed by the delay amount td by the delay circuit 31 to be the signal at the point c. It becomes a waveform. The signal waveforms at points a, b, and c are respectively multiplied by 1/4, 1/2, and 1/4 by multipliers 32, 33, and 34, and added by adder 35 to form a signal at point d. . The signal waveform at the point d passes through the coring circuit 18 to become a signal waveform at the point e, and is further added to the signal waveform at the point b by the adder 36 to become an output signal indicated by the signal waveform at the point f.

  At this time, the coring circuit 18 does not attenuate the frequency characteristic for a minute amplitude signal (for example, 1 to 3 LSB / 8 bits = 256 gradations), so the S / N of the signal is not improved. Is maintained and the image conversion unit 20 increases the definition, so that the signal details can be enhanced.

  On the other hand, although the minute noise is also maintained by the coring circuit 18, the random component in the minute noise remains unemphasized by the image conversion unit 20. This random noise has a randomness that makes it difficult to see false contours in display displays with a problem in the bit accuracy of the display gradation such as a plasma display, so it works as an error diffusion signal and improves the noise feeling. Can do.

  The best mode for carrying out the present invention has been described through the first to third embodiments. However, the present invention is not limited to the above-described embodiments, and departs from the gist of the present invention. It goes without saying that various modifications can be made without departing from the scope.

  According to the present invention described above, an SD image can be converted into an arbitrary high resolution HD image while preventing problems of S / N and preshoot / overshoot.

1 is a block diagram illustrating a schematic configuration of an image processing apparatus according to a first embodiment. It is a block diagram which shows the example of an internal structure of the image conversion part in the image processing apparatus. It is a figure for demonstrating the formation method of a class tap and a prediction tap. It is a figure for demonstrating the formation method of a class tap and a prediction tap. It is a figure for demonstrating the formation method of a class tap and a prediction tap. It is a figure for demonstrating the formation method of a class tap and a prediction tap. It is a block diagram which shows the example of an internal structure of the learning apparatus which learns the prediction coefficient memorize | stored in coefficient ROM of the image conversion part. It is a schematic diagram which shows the frequency characteristic in the signal before and behind the conventional image conversion part. It is a schematic diagram which shows the frequency characteristic in the signal before and behind the image conversion part in this Embodiment, The same figure (A) shows the example in case an image process part has an ideal frequency characteristic, The figure (B) ) Shows an example where the image processing unit has practical frequency characteristics. It is a figure which shows the signal component replaced by the image conversion part. It is a block diagram which shows schematic structure of the image processing apparatus in 2nd Embodiment. It is a block diagram which shows schematic structure of the image processing apparatus in 3rd Embodiment. It is a figure which shows the input / output characteristic of the coring circuit incorporated in the frequency adjustment circuit in the image processing apparatus. It is a figure explaining the frequency adjustment circuit, the figure (A) shows the example of an internal structure of a frequency adjustment circuit, and the figure (B) shows the signal waveform in each point of the figure (A).

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Image processing apparatus, 10 tuner, 11 Image processing part, 12 Y / C separation circuit, 13 Chroma demodulation circuit, 14 Selector, 15 Frequency adjustment circuit, 16 Amplitude level adjustment circuit, 17 S / N detection circuit, 18 Coring circuit 20 image conversion unit, 21 pixel number conversion unit, 22 Y system processing unit, 23 C system processing unit, 24 matrix circuit, 25 drive / cutoff adjustment unit, 26 display display, 27 microcomputer

Claims (12)

Frequency characteristic adjusting means for adjusting the frequency characteristics of the plurality of input first image signals;
Image conversion means for converting the first image signal, the frequency characteristic of which has been adjusted by the frequency characteristic adjustment means, into a second image signal having a larger number of pixels than the first image signal;
The frequency characteristic adjusting unit attenuates the frequency characteristic of the first image signal as a high frequency so as to exhibit a characteristic of a substantially low-pass filter that is substantially the reverse characteristic of the frequency characteristic change by the image converting unit. An image processing apparatus characterized in that , among a plurality of types of first image signals, an image signal having a narrow band and a small amount of high-frequency component information increases the attenuation amount of the frequency characteristic . The first image signal is an S signal, a YUV signal, a CVBS signal, or an RF signal,
The frequency characteristic adjusting means, the S signal and the YUV signal, the CVBS signal, the order of the RF signal, the image processing apparatus according to claim 1, wherein increasing the attenuation amount of the frequency characteristic. The image conversion means includes
Class classification means for classifying a target pixel, which is a pixel of the second image signal of interest, into a predetermined class according to the property of the first image signal corresponding to the target pixel; ,
The image processing apparatus according to claim 1, further comprising: a predicting unit that predicts the predicted value of the target pixel corresponding to the class of the target pixel. The prediction means is
Prediction coefficient storage means for storing, for each class, a prediction coefficient for calculating a prediction value of the target pixel by linear combination with the pixels of the first image signal;
The image processing apparatus according to claim 3, further comprising: a predicted value calculation unit that calculates a predicted value of the target pixel from the prediction coefficient for the target pixel and the pixel of the first image signal . S / N detecting means for detecting S / N of the first image signal is further provided,
The frequency characteristic adjusting means, the result of detection by the S / N detection means, the image processing apparatus according to claim 1, wherein increasing the attenuation amount of the frequency characteristics as S / N is small.   2. The image processing apparatus according to claim 1, wherein the frequency characteristic adjusting unit adjusts the frequency characteristic only in a portion of the first image signal having an amplitude equal to or larger than a predetermined threshold. A frequency characteristic adjustment step of adjusting the frequency characteristics of the plurality of types of input first image signals;
An image conversion step of converting the first image signal whose frequency characteristic has been changed in the frequency characteristic adjustment step into a second image signal having more pixels than the first image signal,
In the frequency characteristic adjustment step, the frequency characteristic of the first image signal is attenuated as a high frequency range so as to show a characteristic of a low-pass filter that is substantially the reverse characteristic of the frequency characteristic change in the image conversion process , An image processing method characterized in that , among a plurality of types of first image signals, an image signal with a narrower band and a smaller amount of high-frequency component information increases the attenuation amount of frequency characteristics . The first image signal is an S signal, a YUV signal, a CVBS signal, or an RF signal,
The image processing method according to claim 7, wherein in the frequency characteristic adjustment step, the attenuation amount of the frequency characteristic is increased in the order of the S signal, the YUV signal, the CVBS signal, and the RF signal. The image conversion step
A class classification step for classifying a target pixel, which is a pixel of the second image signal of interest, into a predetermined class according to the nature of the first image signal corresponding to the target pixel; ,
The image processing method according to claim 7, further comprising: a prediction step of predicting the predicted value of the target pixel corresponding to the class of the target pixel. The prediction process is
A prediction coefficient for calculating a prediction value of the pixel of interest by linear combination with a pixel of the first image signal, stored in a prediction coefficient storage unit for each class, and a pixel of the first image signal ; An image processing method according to claim 9, further comprising: a predicted value calculation step for obtaining a predicted value of the target pixel. An S / N detection step of detecting an S / N of the first image signal;
The image processing method according to claim 7 , wherein, in the frequency characteristic adjustment step, the attenuation amount of the frequency characteristic is increased as the S / N is smaller as a result of the detection in the S / N detection step. 8. The image processing method according to claim 7 , wherein, in the frequency characteristic adjusting step, the frequency characteristic is adjusted only in a portion having an amplitude equal to or larger than a predetermined threshold in the first image signal.

Images