JP2009100171A - Image processor, image display device, and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor capable of reducing quantization noise even in image signals continuously transmitted temporally, an image display device, and an image processing method. <P>SOLUTION: An image signal B output from a gradation correction part 3 is input in a variation calculation part and a delay part of a gradation processing part 4. The variation calculation part calculates difference between the adjacent pixels of the input image signal B as variation G and outputs it to a variation limiting part. The variation limiting part generates variation H by limiting a value so that it is settled within a range indicated by threshold information F output from a gradation correction amount calculation part in the input variation G. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing device, an image display device, and an image processing method.

  Display devices that display images by modulating matrix pixels discretely arranged by light reflection, transmission, or light emission such as liquid crystal, plasma, EL (electroluminescence) and DMD (digital mirror device) In addition to flat-screen televisions and projection televisions, they are used in various image display devices such as projectors and computer monitors. In recent years, high-definition broadcasts and computer processing speeds have increased, and display resolution has been increasing, and device improvements have led to a rapid increase in brightness. With the increase in brightness of displays, quantization noise and gradation jumps when displaying digitized image signals have become conspicuous.

  When an analog image signal is received, quantization noise is reduced by increasing the number of bits of the analog-digital converter (for example, 8 bits to 10 bits). At present, the speed of analog-to-digital converters and the increase in the number of quantizations are significantly increased. In the case of digital transmission, the quantization number cannot be unilaterally changed by agreement between the transmission side and the reception side. Similarly, since the number of quantizations for digital contents on a DVD or network is determined in advance, it is difficult to increase the number of quantizations of the contents themselves because it is necessary to provide a new standard.

  Furthermore, in signal processing for televisions and the like, there are processing for expanding the dynamic range of received image signals to display high-impact video, and processing for converting to gradation-brightness conversion characteristics to convert to high-contrast video. Has been done. The dynamic range expansion process and the change of the gradation-luminance conversion characteristics display a high-contrast video while increasing the quantization noise. Therefore, there is a problem that a pseudo contour is perceived in a continuous gradation change portion that is not a contour in the displayed video or image.

  In order to solve such a problem, a method for reducing quantization noise of an image signal by digital image processing is disclosed in Patent Document 1 below. In Patent Document 1, a combination of gradations in which a gradation jump occurs in accordance with image processing is searched, a region having the combination of gradations is specified from an image signal, and a gradation is determined for the region of the image. A method of reducing quantization noise by performing gradation interpolation processing on neighboring pixels where jumps occur has been disclosed.

  Also, by increasing the number of quantization bits, a combination of gradations where gradation jumps occur is searched, an area having the combination of gradations is identified from the image signal, and the area of the image is scaled. A method of reducing quantization noise by performing gradation interpolation processing on neighboring pixels in which a tone jump has occurred is disclosed.

Japanese Patent No. 3455397 (paragraph 0026, FIG. 1)

  In the gradation interpolation process used to reduce the quantization noise that increases due to the gradation jump that occurs during image processing, as shown in Patent Document 1, the combination of gradations that cause a gradation jump is used. A search is performed in advance, and an area adjacent to the combination of gradations where the gradation jump occurs is specified from the image signal, and gradation interpolation processing is performed around the outline of the specified area. Therefore, since processing cannot be performed in real time, there has been a problem that quantization noise cannot be reduced in an image signal sent continuously in time.

  The present invention has been made in view of the above problems, and an image processing device, an image display device, and an image processing capable of reducing quantization noise even in an image signal sent continuously in time. It aims to provide a method.

  An image processing apparatus according to the present invention calculates a change amount from a gradation correction unit that generates a corrected image signal by performing gradation correction on an original image signal, and a spatial direction difference between pixels of the corrected image signal. A change amount calculation unit, a change amount restriction unit that restricts a possible range of the change amount calculated by the change amount calculation unit, and the corrected image signal from the change amount restricted by the change amount restriction unit An inclination calculation unit that calculates the gradient of the gray scale, an additional bit generation unit that generates an additional bit based on the gradient of the gradation calculated by the inclination calculation unit, and an additional bit generation unit A bit addition unit that generates a processed image signal by adding the additional bit to the corrected image signal, and the gradation correction unit generates threshold information according to the original image signal, and the change The amount limiting unit is configured to use the threshold information. Based on, and determines the limit value that limits the amount of change.

  The image processing apparatus according to the present invention calculates the amount of change from the difference in the spatial direction of the pixels, and performs image signal processing such that the range that the amount of change can take is limited to a threshold value or less. Therefore, quantization noise can be reduced even in image signals sent continuously in time.

<Embodiment 1>
FIG. 1 is a block diagram showing a configuration of an image display apparatus according to Embodiment 1 of the present invention. The image display apparatus shown in FIG. 1 displays an image having brightness according to gradation data constituting pixels of a received image signal. The outline of the operation of the image display apparatus will be described below with reference to FIG.

  An analog or digital image signal is input to the input terminal 1 from the outside. The input of the image signal to the input terminal 1 may be an image signal transmitted using an electrical cable, or an image signal transmitted using radio waves or light. . The image signal input to the input terminal 1 is received by the receiving unit 2 and converted into an image signal format used for subsequent processing. For example, when the image signal input to the input terminal 1 is an analog signal, it is converted into a digital image signal by an analog-digital converter, and when it is a serial digital image signal, a predetermined image signal is converted by serial-parallel conversion. Processing that matches the transmission method is performed. The image signal has a format represented by luminance and chromaticity and a format represented by primary color signals such as red, green, and blue. The gradation processing unit 4 processes each signal in the same manner. can do.

  The image signal A output from the reception unit 2 is input to the gradation correction unit 3. The gradation correction unit 3 performs gradation correction (gradation conversion) according to the gradation distribution of the input image signal.

  FIG. 2 is a block diagram showing a configuration of the gradation correction unit 3. The image signal A input from the reception unit 2 is input to the gradation distribution detection unit 6 and the gradation conversion unit 8. The gradation distribution detection unit 6 obtains the frequency of gradation every frame of the image signal A or every several frames. The gradation distribution detection unit 6 outputs the detected gradation distribution information D to the gradation correction amount calculation unit 7. The gradation correction amount calculation unit 7 determines the gradation correction amount based on the gradation distribution information D detected by the gradation distribution detection unit 6 and outputs the gradation correction parameter E to the gradation conversion unit 8. The threshold information F is output to the gradation processing unit 4 at the subsequent stage.

  FIG. 3 is a graph showing the operation of the gradation correction unit 3. FIG. 3A is an example of the gradation distribution detected by the gradation distribution detection unit 6. FIG. 3A shows a case where neither black nor white pixels exist in the pixels of the image signal A. The gradation distribution detection unit 6 outputs this gradation distribution to the gradation correction amount calculation unit 7 as gradation distribution information D. Based on the gradation distribution information D, the gradation correction amount calculation unit 7 calculates the cumulative frequency from the minimum gradation (black) and the cumulative frequency from the maximum gradation (white), and compares it with a predetermined threshold value. Then, the black side (input gradation) value B1 and the white side (input gradation) value W1 shown in FIG. 3B are obtained. Further, the gradation correction amount calculation unit 7 determines the gradation correction amount so that the value B1 and the value W1 are output as the (output gradation) value B2 and the (output gradation) value W2, for example, FIG. A gradation correction parameter E composed of a coefficient of an arithmetic expression that realizes the conversion characteristic of 3 (b) and data of a lookup table is output to the gradation conversion unit 8.

  FIG. 3C is a diagram showing a frequency distribution as a result of the tone conversion performed by the tone converting unit 8 on the basis of the tone correction parameter E of the image signal A input from the receiving unit 2. As shown in the figure, the gradation conversion generates pixels of the minimum gradation (black) and the maximum gradation (white), the dynamic range of the image is expanded, and an image with high contrast can be displayed. The tone conversion unit 8 outputs the tone-converted image signal B to the tone processing unit 4.

  The broken line shown in FIG. 3B shows the input / output characteristics of the image when gradation conversion is not performed. As can be seen from this figure, in the gradation conversion characteristics in the gradation converting unit 8, the gradient from the minimum gradation to the value B1 and the gradient from the value W1 to the maximum gradation are smaller than 1, and the values B1 and W1 The slope between is greater than 1. That is, in the image signal A, a gradation jump occurs at a gradation between the value B1 and the value W1. The gradation correction amount calculation unit 7 distinguishes between the region from the minimum gradation to the value B1, the region from the value W1 to the maximum gradation, and the region from the value B1 to the value W1, and is suitable for each region. The threshold information F is output to the gradation processing unit 4.

  Note that the image used by the gradation distribution detection unit 6 to create the gradation distribution may be an entire screen or a part of the screen. For example, when there are black bands above and below in an image signal of a cinema or the like, optimum gradation correction can be performed by detecting the gradation distribution in an area excluding the black band portion. Further, when the image display device itself displays a menu or the like on a part of the screen and the menu image is superimposed on the image signal A, the menu part is removed from the gradation distribution detection area and the gradation is By not performing the correction, the influence of the menu image on the detection of the gradation distribution can be suppressed, and the brightness of the menu or the like can be made not to change due to the gradation distribution of the image signal A.

  FIG. 4 is a block diagram showing a configuration of the gradation processing unit 4. The image signal B output from the gradation correction unit 3 is input to the change amount calculation unit 9 and the delay unit 13 of the gradation processing unit 4. The change amount calculation unit 9 calculates a difference between adjacent pixels of the input image signal B as a change amount G and outputs the difference to the change amount limiting unit 10. The change amount limiting unit 10 generates the change amount H by limiting the value of the input change amount G so that it falls within the range indicated by the threshold information F output from the gradation correction amount calculation unit 7. The change amount restriction unit 10 outputs the change amount H generated by restricting the value to the inclination determination unit 11.

  The inclination determination unit 11 determines the gradient of the change in gradation in one direction of the image signal B using the change amount H input from the change amount limiting unit 10, and as the determination result, the gradient inclination I and the inclination The count value N obtained inside the determination unit 11 is output to the additional bit generation unit 12. The count value N will be described later.

  The additional bit generation unit 12 obtains the amount of change between adjacent pixels from the input gradation gradient I and the count value N, and outputs it to the bit addition unit 14 as an additional bit J. The change amount obtained here is a change amount including a decimal value with respect to the gradation value of the image signal B.

  On the other hand, the delay unit 13 delays the image signal B output from the gradation conversion unit 8 by a predetermined time, so that the additional bit J is correctly added to the image signal B by the bit addition unit 14 at the subsequent stage. Thus, the image signal K obtained by delaying the image signal B is generated and output.

  The image signal K delayed by a predetermined time by the delay unit 13 is input to the bit addition unit 14. The bit addition unit 14 adds the additional bit J output from the additional bit generation unit 12 to the image signal K output from the delay unit 13 to generate and output a new image signal C. The image signal C output from the bit addition unit 14 is input to the display unit 5 and an image is displayed.

  Hereinafter, the operation of the gradation processing unit 4 will be described in detail with reference to FIG.

  FIG. 5A shows a part of the pixels arranged in the horizontal direction of the image signal B output from the gradation correction unit 3. The vertical direction indicates the gradation, and the horizontal direction indicates the horizontal pixel position. It is assumed that the leftmost step indicates a difference of one gradation of the image signal B. Here, the case where the image signal B is composed of 10 bits will be described. However, the present invention is not limited to this. The illustrated gradation is the minimum gradation width (resolution) represented by the number of bits constituting the image signal. ).

  FIG. 5B shows the difference between adjacent pixels of the image signal B calculated by the change amount calculation unit 9, that is, the change amount G. When the gradation value of the pixel at the horizontal position x is P (x), the difference between the pixels at the horizontal position x can be expressed by P (x) −P (x−1). The change amount G is output from the change amount calculation unit 9 to the change amount restriction unit 10.

  FIG. 5C shows the change amount G, that is, the change amount H whose value is limited by the change amount limiting unit 10 based on the threshold information F. Here, a case where the threshold information F is 1 will be described. When the threshold information F is 1, the change amount restriction unit 10 restricts the value of the change amount G so that it falls within a range from −1 to +1. That is, the amount of change H output from the amount-of-change limiting unit 10 includes three values of +1, 0, and −1. Here, of the change amount G, the fourth change amount from the left and the rightmost change amount that are outside the range from −1 to +1 are limited to +1 and −1, respectively.

  FIG. 5 (d) shows the concept of the operation of the inclination determination unit 11. First, the inclination determination unit 11 starts from the change amount H input from the change amount restriction unit 10 and the number of zeros (pixels). Number), that is, the width of the pixel whose gradation has not changed. Next, a fine gradient I of the gradation is calculated from the detected width (distance) and the amount of change H, and is output to the additional bit generation unit 12. For example, when the change amount H of 7 pixels is continuous for 7 pixels, the distance in which the gradation has not changed is 8. If the value of the subsequent change amount H is 1, then the gradation change slope I in that section is 1. Becomes 1/8. Further, when 0 of the change amount H is continuous for 10 pixels, the distance is 11, and when the value of the change amount H after that is −1, the gradient I of gradation change in that section is −1 / 11

  Referring to FIG. 5E, the additional bit generation unit 12 includes a decimal part for the gradation value of the image signal B based on the gradient I and the count value N input from the gradient determination unit 11. An additional bit J consisting of a bit string is generated and output to the bit adder 14. The operation of generating the additional bit J will be described in detail later with reference to other drawings.

  Referring to FIG. 5F, the additional bit J obtained by the additional bit generation unit 12 is output to the bit addition unit 14, and the bit addition unit 14 includes the fractional part output from the additional bit generation unit 12. The additional bit J and the image signal K output from the delay unit 13 are added to generate the image signal C. Since the additional bit J includes a decimal part with respect to the gradation value of the image signal K (or image signal B), the added image signal C has an accuracy with a decimal part with respect to the image signal K. Image signal. That is, since the number of bits of the image signal C is larger than the number of bits of the image signal K, the number of gradations that can be expressed can be increased. Therefore, the gradation resolution is increased, and a smoother gradation change can be expressed. The image signal C having the increased number of bits is output to the display unit 5 and displayed.

  FIG. 6 shows a part of the image signal B, and the detailed operation of the gradation processing unit 4 will be described with reference to FIG.

  FIG. 6 shows a part of the image signal B composed of three gradations. In the image signal B, 8 pixels with gradation value 1, 8 pixels with gradation value 2, and 1 pixel with gradation value 3 are arranged in order from the left pixel. That is, if the image signal B is composed of 10 bits, a part of the image that is brightened by one gradation from left to right is shown with respect to the 10-bit image signal.

  The change amount calculation unit 9 obtains a difference between adjacent pixels as a change amount G with respect to the arrangement of gradation values of the image signal B. In FIG. 6, the change amount is +1 for the pixel with the gradation value 2 adjacent to the gradation value 1 and the pixel with the gradation value 3 adjacent to the gradation value 2, and the change amount is 0 for the other pixels. . This change amount G is output to the change amount limiting unit 10.

  When the threshold information F is 1, the change amount restriction unit 10 restricts the value of the change amount G so that the change amount G falls within the range from −1 to +1. The change amount H whose value is limited is output to the inclination determination unit 11. In FIG. 6, since the change amount G is composed of values of 0 and +1, the value is not limited. However, when the change amount G is greater than 1, the change amount H is limited to 1. When the change amount G is smaller than -1, the change amount H is limited to -1. This limited range is determined based on the threshold information F.

  The inclination determination unit 11 is configured by a counter or the like, and counts the number of consecutive 0s of the change amount H whose value is limited. That is, the number of changes H is counted, and when it is other than 0, the count value at that time is stored, and the value of the counter is returned to 0. In FIG. 6, the count value N returns to 0 in the leftmost pixel, and thereafter, counting is performed until the change amount H is other than 0, that is, +1. Here, the count value N is counted up to 7, then returns to 0, and the count starts again. This count value N = 7 indicates that the number of pixels in the section where 0 is continuous (the section where the gradation value of the image signal B is 1) is 8, and the inclination determination unit 11 The gradation gradient I is determined as 1/8. Similarly, since the count value N is 7 in the interval where the gradation value B is 2, the gradation gradient I in the interval is also determined to be 1/8. The inclination determination unit 11 outputs the inclination I and the count value N to the additional bit generation unit 12 as a determination result.

  The additional bit generation unit 12 obtains a gradation value to be added to the image signal B (image signal K) using the inclination I and the count value N included in the determination result input from the inclination determination unit 11. The bit string indicating the gradation value is output to the bit adder 14 as an additional bit J. In FIG. 6, the gradation value represented by the additional bit J to be added to the image signal B is shown as the additional value J. Here, since the minimum gradation of the image signal B is represented as 1, the additional value J is represented as a decimal number of 1 or less.

  The additional value J can be expressed by J = N * I from the count value N and the slope I. In FIG. 6, a sequence of additional values J having a change of the obtained slope I = 1/8 is obtained, such as 0/8, 1/8, 2/8... From the left.

  By adding this additional value J to the image signal K by the bit adder 14, a gradation value having an accuracy of 1 or less can be obtained as indicated by the gradation value KJ after the addition.

  5 and 6, the case where the threshold information F is 1 has been described, but the threshold information F is determined by the gradation correction amount calculation unit 7 according to the gradation conversion characteristics of the gradation conversion unit 8. As described above, in the gradation conversion characteristics shown in FIG. 3, the threshold information F is small for the gradation from the minimum gradation to the value B1 and the gradation from the value W1 to the maximum gradation, and from the value B1 to the value. The threshold information F is set large for each of the gradations up to W1. In the following, the threshold information F for the gradation from the minimum gradation to the value B1 and the gradation from the value W1 to the maximum gradation is 1, and the threshold information F for the gradation from the value B1 to the value W1 is 2. Will be described.

  FIG. 7 (and FIG. 8) is a diagram for explaining the operation when the threshold information F is different depending on the gradation, unlike FIG. 5 (and FIG. 6) in which the threshold information F is fixed to 1. is there.

  In the image signal B of FIG. 7A, the interval Ba indicates a value when the gradation of the image signal B is from the minimum gradation to the value B1, and the interval Bb indicates the gradation of the image signal B from the value B1. The values up to the value W1 are shown. In the section Ba, the image signal B changes with a minimum gradation width of 1, but in the section Bb, the minimum gradation width of the image signal B changes with 2 by the processing of the gradation correction unit 3 is shown. In the actual process, the interval Bb may include a case where the minimum gradation width is 1 and a case where it is 2, but here, all will be described as 2.

  FIG. 7B shows the difference between adjacent pixels of the image signal B calculated by the change amount calculation unit 9, that is, the change amount G. In the interval Ba, the value of the change amount G is +1, at the boundary between the interval Ba and the interval Bb, the value of the change amount G is greater than 2, and in the interval Bb, the value of the change amount G is +2 or -2. Appear.

  In FIG. 5C, the change amount H whose value is limited by the change amount limiting unit 10 based on the threshold information F is shown, but in FIG. 7C, the threshold information F of the section Ba is represented by the value SH1. The amount of change H is expressed using the threshold information F of the section Bb as the value SH2. Since SH1 = 1 in the interval Ba, the change amount limiting unit 10 limits the value of the change amount G so that it falls within the range from −1 to +1. That is, the change amount G output from the change amount limiting unit 10 is composed of three values of +1, 0, and −1. Further, since SH2 = 2 in the section Bb, the change amount limiting unit 10 limits the value of the change amount G so that it falls within the range from −2 to +2. That is, the amount of change G output from the amount-of-change limiting unit 10 is composed of five values of +2, +1, 0, -1, and -2. Here, in the change amount G, there is a change exceeding SH1 at the boundary portion from the section Ba to the section Bb and the boundary portion from the section Bb to the section Ba, and each is limited to −1 and +1. It is shown.

  The operations in FIGS. 7D to 7F are the same as those in FIGS. 5D to 5F, and detailed description thereof is omitted.

  FIG. 8 is a diagram for explaining a more detailed operation of the gradation processing unit 4 in the gradation corresponding to the section Bb in FIG. The operation in the section Ba is the same as that described in FIG.

  FIG. 8 shows a part of the image signal B composed of three gradations. In the image signal B, 8 pixels with a gradation value of 80, 8 pixels with a gradation value of 82, and 1 pixel of a gradation value of 84 are arranged in order from the left pixel. That is, if the image signal B is composed of 10 bits, a part of the image that is brightened by two gradations from left to right is shown with respect to the 10-bit image signal. This shows a case where the minimum gradation width has become 2 by the processing of the gradation correction unit 3.

  The change amount calculation unit 9 obtains a difference between adjacent pixels as a change amount G with respect to the arrangement of gradation values of the image signal B. In FIG. 8, the change amount G is +2 in the pixel of the gradation value 82 adjacent to the gradation value 80 and the pixel of the gradation value 84 adjacent to the gradation value 82, and the change amount G is 0 in the other pixels. It becomes. This change amount G is output to the change amount limiting unit 10.

  When the SH2 of the threshold information F is 2, the change amount restriction unit 10 restricts the value of the change amount G so that the change amount G falls within a range from −2 to +2. The change amount H whose value is limited is output to the inclination determination unit 11. In FIG. 8, since the change amount G is composed of values of 0 and +2, the value is not limited. However, when the change amount G is greater than 2, the change amount H is limited to 2. When the change amount G is smaller than −2, the change amount H is limited to −2. This limited range is determined based on the threshold information F.

  The inclination determination unit 11 is configured by a counter or the like, and counts the number of consecutive 0s of the change amount H whose value is limited. That is, the number of changes H is counted, and when it is other than 0, the count value at that time is stored, and the value of the counter is returned to 0. In FIG. 8, the count value N returns to 0 in the leftmost pixel, and thereafter, counting is performed until the change amount H is other than 0, that is, +2. Here, the count value N is counted up to 7, then returns to 0, and the count starts again. This count value N = 7 indicates that the number of pixels in the section where 0 is continuous (the section where the gradation value of the image signal B is 80) is 8, and the inclination determination unit 11 The gradient I is determined as 2/8. Similarly, since the count value N is 7 in the section where the gradation value B is 82, the gradation gradient I in the section is also determined to be 2/8. The inclination determination unit 11 outputs the inclination I and the count value N to the additional bit generation unit 12 as a determination result.

  The additional bit generation unit 12 obtains a gradation value to be added to the image signal B (image signal K) using the inclination I and the count value N included in the determination result input from the inclination determination unit 11. The bit string indicating the gradation value is output to the bit adder 14 as an additional bit J. In FIG. 8, the gradation value represented by the additional bit J to be added to the image signal B is shown as the additional value J. Here, since the minimum gradation of the image signal B is represented as 2, the additional value J is represented as (decimal number of 1 or less + 1).

  The additional value J can be expressed by J = N * I from the count value N and the slope I. In FIG. 8, a sequence of additional values J having a change of the obtained gradient I = 2/8 is obtained, such as 0/8, 2/8, 4/8... From the left.

  By adding this additional value J to the image signal K by the bit adder 14, a gradation value having an accuracy of 1 or less can be obtained as indicated by the gradation value KJ after the addition.

  FIG. 9 is a diagram illustrating the relationship between the additional bit J output from the additional bit generation unit 12, the bit of the image signal K output from the delay unit 13, and the bit of the image signal C output from the bit addition unit 14. . Here, a case where the image signal K is 10 bits, the additional bit J is 6 bits (including a sign bit), and the image signal C is 13 bits will be described. As shown in FIG. 8, the additional bit J includes a sign, an integer part, and a decimal part with respect to the image signal K. Therefore, the bit addition unit 14 matches the decimal point of the image signal K with the decimal point of the additional bit J. Add to. At this time, the sign bit S of the additional bit J is used for the upper bits of the image signal K. Since the image signal C has a bit string extending downward with respect to the image signal K, that is, it includes a decimal part with respect to the original image signal, the resolution of one gradation of the image signal C is higher than that of the image signal K. Therefore, it becomes possible to express a gradation (high accuracy) with less quantization noise. When the additional bit J having the sign bit is added to the image signal K, the sign bit is repeatedly used so that the number of bits of the integer part of the additional bit J is the same as the integer part of the image signal K. Needless to say.

  FIG. 10 is a block diagram showing another configuration of the gradation correction unit 3. The image signal A output from the receiving unit 2 is input to the delay unit 15, delayed by a predetermined period by the delay unit 15, and then input to the gradation correction unit 8 as the image signal L. The gradation distribution detection operation in the gradation distribution detector 6 requires at least one frame of image data, and as a result, a delay of one frame with respect to the image signal A occurs. By providing the delay unit 15 with a delay generated by the gradation distribution detection unit 6 and the gradation correction amount calculation unit 7, the timing of the image signal and the gradation correction can be matched. By this delay operation, it is possible to suppress blinking of an image due to a delay in gradation correction. Other operations are the same as those in FIG.

  FIG. 11 is a graph showing an example of another operation of the gradation correction unit 3. In the description using FIG. 3, the case where the gradation conversion characteristic is obtained by the cumulative frequency of the gradation and the threshold value has been described. However, the present invention is not limited to this, and gradation conversion is performed from the gradation distribution detected by the gradation distribution detection unit 9. The characteristics can also be determined. FIG. 11A is a graph showing the gradation distribution of the image signal B. FIG. The gradation distribution D divides the input gradation into a plurality of areas, and counts the number of pixels belonging to each area. As an example, there is a case where an area is divided into 32 for a 10-bit image signal. The gradation distribution D obtained here is output to the gradation correction amount calculation unit 7. Based on this gradation distribution, the gradation correction amount calculation unit 7 increases the gradient of the gradation conversion characteristic in the high-frequency area and decreases the gradient of the gradation conversion characteristic in the low-frequency area. Determine key conversion characteristics. In FIG. 11B, since the frequency of the region g1 is approximately 0, the slope is smaller than 1, the slope of the region g2 is about 1, and the region g3 is set to a value larger than 1. By performing gradation correction in this way, an image with a sense of contrast can be displayed. At this time, the threshold value information F output from the gradation correction calculation unit 7 is also set to be small in the regions g1 and g2 and large in the region g3, so that no gradation jump is generated by the gradation processing unit 4. Accordingly, in the areas g1 and g2, the resolution is increased and a smoother image can be displayed. In the area g3, the gradation jump is reduced and the resolution is increased, thereby displaying a smoother image. Will be able to.

  FIG. 12 is a diagram for explaining another operation of the gradation processing unit 4. Here, a function for monitoring the change in the polarity of the change amount H is provided inside the inclination determination unit 11 of the gradation processing unit 4, and when the polarity of the change amount H changes from positive to negative or from negative to positive, Operate so that the slope is zero. In FIG. 12, the polarity of the amount of change H changes from positive to negative in the portion G1, and the slope I of the portion G1 is set to zero. Accordingly, since the additional bit J is also 0 in the portion G1, the gradation of the input image signal is held in the image signal C output from the bit addition unit 14 to suppress the occurrence of unnecessary blurring. Can do.

  In the description of the above operation, the gradation processing unit 3 reduces quantization noise by adding additional bits generated according to the amount of change in gradation in the horizontal direction of the input image signal to the image signal. However, the present invention is not limited to this, and the gradation processing unit 3 generates additional bits according to the amount of change in gradation in the vertical direction of the input image signal and adds it to the image signal. The same effect can be obtained.

  As described above, the generation of the additional bit J mainly involves increasing the number of bits of data indicating the slope obtained from the counting result in the slope determination unit 11, so that it is higher without increasing the circuit scale. An additional bit of accuracy can be obtained. Therefore, it is easy to generate more image data than the number of bits that can be input to the display unit 5. In this case, the gradation processing unit 4 once generates image data larger than the number of bits of the display unit 5, and the bit number of the image data increased by error diffusion, dithering, or the like is changed to the number of bits of the display unit 5. By combining them, it is possible to display an image having a larger number of gradations than the number of gradations that the display unit 5 has.

  As described above, the image processing apparatus according to the present embodiment includes the gradation correction unit 3 that generates the image signal B by performing gradation correction on the original image signal input from the outside, and the pixels of the image signal B. The change amount calculation unit 9 that calculates the change amount from the difference in the spatial direction, the change amount restriction unit 10 that restricts the range that the change amount G calculated by the change amount calculation unit 9 can take, and the change amount restriction unit 10 restrict The gradient determination unit 11 that calculates the gradient I of the gradation of the image signal B from the change amount H, and the additional bit generation that generates the additional bit J based on the gradient I calculated by the gradient determination unit 11 Unit 12 and a bit addition unit 14 that generates an image signal C by adding the additional bit J generated by the additional bit generation unit 12 to the image signal B. The gradation correction unit 3 includes the original image signal Threshold value information F corresponding to the Based on the threshold value information F, to determine the limit value that limits the amount of change G. Therefore, as compared with Patent Document 1, it is possible to reduce the quantization noise even in an image signal sent continuously in time.

  In the above description of the operation, the case where the gradation processing unit 4 reduces the gradation jump generated in the gradation correction unit 3 has been described. However, the generation of the gradation jump is only the gradation correction that improves the contrast. The tone correction unit 4 is arranged behind the image processing that generates the tone jump, and the tone jump that occurs in the image processing can be reduced. it can.

  The above operation is not limited to one direction in the horizontal direction or the vertical direction, and may be sequentially performed in both the horizontal direction and the vertical direction.

<Embodiment 2>
In the first embodiment, as shown in FIG. 1, the amount of change G is limited using the threshold information F generated in the gradation correction unit 3 for the image data received from the outside by the reception unit 2. Explained when to do. However, the image data does not necessarily have to be received from the outside by the receiving unit 2, and the threshold information F does not necessarily have to be generated by the gradation correcting unit 3.

  FIG. 13 is a block diagram showing a configuration of an image display apparatus according to Embodiment 2 of the present invention.

  The input terminal 16 is a terminal for inputting encoded image data, that is, encoded image data. Examples of encoded image data include encoded image data received by a tuner in digital broadcasting, encoded image data input via a network such as the Internet, and a storage medium such as a DVD, a hard disk, or a semiconductor memory. The read encoded image data can be mentioned.

  The encoded image data input from the input terminal 16 is input to the decoding unit 17 and decoded, thereby generating an image signal A. The decoding unit 17 includes a decoding unit and an encoding error detection unit (not shown). The decoding unit 17 generates and outputs the image signal A at the same time, and at the same time, the encoding error detection unit includes a quantum included in the decoded image. The amount of quantization error (or encoding error) is detected, and threshold information F is output based on the quantization error.

  At this time, the threshold information F is set to a large value when the quantization error of the decoded image is large, and is set to a small value when the quantization error is small. The threshold information F is input to the gradation processing unit 4. In the gradation processing unit 4, as in the operation of the first embodiment, the quantization noise and gradation of the image signal A output from the decoding unit 17 based on the threshold information F input from the encoding unit 17. Processes to reduce jumps.

  By configuring as described above, it is possible to reduce the quantization error that occurs when an image whose information amount is compressed by encoding the image is decoded, and there is less quantization noise (high accuracy). Gradation can be expressed.

  As described above, in the image processing device, the image display device, and the image processing method according to the present embodiment, the encoded image data is generated based on the quantization error in the encoding error detection unit of the decoding unit 17. Using the threshold information F, the image signal is processed so that the range of change G can be limited to the threshold or less, as in the first embodiment. Therefore, even when the decoding unit 17 performs dynamic range expansion or contrast correction processing, the change amount limiting values (maximum value, minimum value) in the change amount limiting unit 10 are the characteristics of the decoding unit 17. By setting them together, it is possible to perform processing without searching for the gradation at which gradation jump occurs, reducing quantization noise and gradation jump in real time and displaying a smooth and high-quality image. can do.

It is a block diagram which shows the structure of the image display apparatus which concerns on Embodiment 1 of this invention. It is a block diagram which shows the structure of a gradation correction | amendment part. It is a graph which shows operation | movement of a gradation correction | amendment part. It is a block diagram which shows the structure of a gradation process part. It is a figure which shows operation | movement of a gradation process part. It is a figure which shows operation | movement of a gradation process part. It is a figure which shows operation | movement of a gradation process part. It is a figure which shows operation | movement of a gradation process part. It is a figure which shows the relationship between the additional bit which an additional bit production | generation part outputs, the bit of the image signal which a delay part outputs, and the bit of the image signal which a bit addition part outputs. It is a block diagram which shows the structure of a gradation correction | amendment part. It is a graph which shows operation | movement of a gradation correction | amendment part. It is a figure which shows operation | movement of a gradation process part. It is a block diagram which shows the structure of the image display apparatus which concerns on Embodiment 2 of this invention.

Explanation of symbols

  1, 16 input terminals, 2 receiving unit, 3 tone correcting unit, 4 tone processing unit, 6 tone distribution detecting unit, 7 tone correcting amount calculating unit, 8 tone converting unit, 9 change calculating unit, 10 Change amount limiting unit, 11 determining unit, 12 additional bit generating unit, 13, 15 delay unit, 14 bit adding unit, 17 decoding unit, A, B, C, K image signal, D gradation distribution information, E gradation Correction parameter, F threshold information, G and H variation, I gradation gradient, J additional bits, N count value.

Claims (5)

A gradation correction unit that generates a corrected image signal by performing gradation correction on the original image signal;
A change amount calculation unit that calculates a change amount from the difference in the spatial direction of the pixels of the corrected image signal;
A variation limiting unit that limits a possible range of the variation calculated by the variation calculating unit;
An inclination calculating unit that calculates an inclination of a gradation of the corrected image signal from the amount of change limited by the amount of change limiting unit;
An additional bit generation unit that generates additional bits based on the gradient of the gradation calculated by the inclination calculation unit;
A bit addition unit that generates a processed image signal by adding the additional bit generated by the additional bit generation unit to the corrected image signal;
The gradation correction unit generates threshold information according to the original image signal,
The change amount restriction unit determines a restriction value for restricting the change amount based on the threshold information. The image processing apparatus according to claim 1,
The additional bit generation unit generates the additional bit having a quantized value corresponding to a fractional part of the corrected image signal;
The image processing apparatus, wherein the bit addition unit generates the processed image signal having a larger number of quantization than the corrected image signal. The image processing apparatus according to claim 1 or 2,
A receiving unit for receiving the original image signal from the outside;
An image display device comprising: a display unit that displays the processed image signal. A change amount calculating step of calculating a change amount from the difference in the spatial direction of the pixel of the pre-processing image signal;
A change amount limiting step of limiting a possible range of the change amount calculated in the change amount calculating step based on gradation conversion characteristics;
An inclination calculating step of calculating a gradient of a gradation of a partial area of the pre-processing image signal from the amount of change limited in the amount of change limiting step;
Generating a post-processing image signal having a quantization error smaller than a quantization error of the pre-processing image signal based on the gradient of the gradation calculated in the tilt calculation step. Processing method. A change amount calculating step of calculating a change amount from the difference in the spatial direction of the pixel of the encoded image signal before processing;
A decoding step of decoding the pre-processed image signal and calculating an encoding error;
A change amount limiting step of limiting a possible range of the change amount calculated in the change amount calculating step based on the encoding error calculated in the decoding step;
An inclination calculating step of calculating a gradient of a gradation of a partial area of the pre-processing image signal from the amount of change limited in the amount of change limiting step;
Generating a post-processing image signal having a quantization error smaller than a quantization error of the pre-processing image signal based on the gradient of the gradation calculated in the tilt calculation step. Processing method.

Images